ARXVision translates visual and spatial information of surroundings into meaningful audio cues through the use of AI, Machine Learning and User-centered design. For the blind and visually impaired. Formerly, Horus Sight.

ARxVision

As AI services and the data they consume and create become more important and prevalent in various applications and processes, so do the platforms and architectures they are built upon. As usual, there is no “one size fits all”, however, what is briefly presented here is an optimized approach to such data-driven AI application architectures. All of the source code mentioned and more can be found , and a free “Develop with Oracle AI and Database Services: Gen, Vision, Speech, Language, and OML” workshop (where all of the use cases are based on the U.N.’s 17 Sustainable Development Goals) giving many more examples can be found . here here Often, multiple network calls must be made in a given AI app, entailing calls to the AI services as well as calls to retrieve and persist the content (whether it be text, audio, images, video, etc.) that is the input or output. The persistent information is then often processed and analyzed further and additional calls, AI or otherwise, are made in reaction. The provides the ability to make calls out to other services, again AI and otherwise, whether they be within the Oracle Cloud or external. Oracle database When the calls are instead made from the database itself, it provides an optimized architecture with various benefits including: Reduced network calls, thus reducing latency. Reduced network calls, thus, increasing reliability. Transactional (ACID) operations on AI and other data (and even messaging when using TxEventQ) which avoid the need for idempotent/duplicate processing logic, etc., and the related wasted resources there. Processing optimization is due to the locality of data whether that data is stored directly in the database or an object store or other source. This is because the Oracle database provides a robust functional frontend to otherwise dumb object storage buckets, and the database provides many options to either sync or optimally operate on data in place in the object store and other data sources. Enhanced security due to a common authentication mechanism and reuse of a famously robust database and cloud security infrastructure. Reduced overall configuration as calls are made from a central location. The entry point to the database itself can be exposed as a Rest endpoint (using ORDS) with a single click, and of course, drivers in various languages can be used to access the database as well. Vector database advantages. This topic is a blog unto itself, and one I’ll release as a follow-on especially as Oracle has and is adding several powerful features in this area. Oracle Database Machine Learning. In addition to various AI services, the Oracle database itself has had a machine-learning engine for many years. OML streamlines the ML lifecycle, offering scalable SQL, R, Python, REST, AutoML, and no-code tools with over 30 in-database algorithms, enhancing data synchronization and security by processing data directly in the database. Oracle Autonomous Database. Select AI which enables querying data using natural language and generating SQL that is specific to your database. Oracle Autonomous Database. AI Vector Search, which includes a new vector data type, vector indexes, and vector search SQL operators, enables the Oracle Database to store the semantic content of documents, images, and other unstructured data as vectors, and use these to run fast similarity queries. These new capabilities also support RAG (Retrieval Augmented Generation) which provides higher accuracy and avoids having to expose private data by including it in the LLM training data. Again, there are many different AI application flows and requirements, but a basic comparison of the two approaches can be visualized in the following way.: The Code It is possible to run several different languages in the database, making it possible to conduct various application logic there. These include Java, JavaScript, and PL/SQL. PL/SQL examples are given here, and they can be executed from the page in the OCI console, from the SQLcl command line tool (which is pre-installed in the OCI Cloud Shell or can be downloaded), from SQLDeveloper, VS Code (where Oracle has a convenient plugin), etc. Database Actions -> SQL There are also a few ways to go about making the calls out to AI and other services. Standard Rest calls using the database’s UTL_HTTP package or fetch from JavaScript, etc., is one approach. If the AI services run within OCI (the Oracle Cloud), then OCI SDKs, which are written for all major languages, can also be used. I find the use of the DBMS_CLOUD.send_request package for all OCI services calls (rather than, for example, more specific OCI SDK calls such as DBMS_CLOUD_OCI_AIV_AI_SERVICE_VISION) to be the simplest and most dynamic approach. We start by creating a credential that can be referenced and reused for all of our cloud service calls and simply includes the information from your OCI account/config. BEGIN
dbms_cloud.create_credential (
 credential_name => 'OCI_KEY_CRED',
 user_ocid => 'ocid1.user.oc1..[youruserocid]',
 tenancy_ocid => 'ocid1.tenancy.oc1..[yourtenancyocid]',
 private_key => '[yourprivatekey - you can read this from file or put the contents of your pem without header, footer, and line wraps]'
 fingerprint => '[7f:yourfingerprint]'
);
END; Next, before we look at the main program/function, let’s just quickly look at the table we’ll save the AI results in. Notice, in this case, the table has columns for both the JSON from an AI call return and a text field that is created from key fields in the JSON for quick reference, searches, etc. Again, the table structures, use of SQL/relational vs JSON, etc., may all vary, and again, this is a great example of the Oracle multi-purpose database where you can use various data models and data types. For example, the JSON Duality feature in the Oracle database is worth checking out as it allows the same data to be accessed using SQL/relational as well as JSON and even MongoDB APIs. CREATE TABLE aivision_results
    (id RAW (16) NOT NULL,
     date_loaded TIMESTAMP WITH TIME ZONE,
     label varchar2(20),
     textfromai varchar2(32767),
     jsondata CLOB
     CONSTRAINT ensure_aivision_results_json CHECK (jsondata IS JSON));
/ And now, the simple function that typifies the heart of the architecture… Here, we see a call to DBMS_CLOUD.send_request with the credential we created and the URL of the (AI) service operation endpoint (the operation of the Oracle Vision AI service in this case). analyzeImage The JSON payload of the body consists of the feature(s) of the service that we would like to use and any other config as well as the arguments to the operation which, in this case, include the object storage location of an image (another option would be to provide the image byte array directly/inlined as part of the payload). The JSON result is then retrieved from the response, certain elements of it are parsed out into a text field for convenience, and the JSON, text, etc., are persisted as mentioned earlier. CREATE OR REPLACE FUNCTION VISIONAI_TEXTDETECTION (
    p_endpoint VARCHAR2,
    p_compartment_ocid VARCHAR2,
    p_namespaceName VARCHAR2,
    p_bucketName VARCHAR2,
    p_objectName VARCHAR2,
    p_featureType VARCHAR2,
    p_label VARCHAR2
) RETURN VARCHAR2 IS
    resp DBMS_CLOUD_TYPES.resp;
    json_response CLOB;
    v_textfromai VARCHAR2(32767);
BEGIN
    resp := DBMS_CLOUD.send_request(
        credential_name => 'OCI_KEY_CRED',
        uri => p_endpoint || '/20220125/actions/analyzeImage',
        method => 'POST',
        body => UTL_RAW.cast_to_raw(
            JSON_OBJECT(
                'features' VALUE JSON_ARRAY(
                    JSON_OBJECT('featureType' VALUE p_featureType)
                ),
                'image' VALUE JSON_OBJECT(
                    'source' VALUE 'OBJECT_STORAGE',
                    'namespaceName' VALUE p_namespaceName,
                    'bucketName' VALUE p_bucketName,
                    'objectName' VALUE p_objectName
                ),
                'compartmentId' VALUE p_compartment_ocid
            )
        )
    );
    json_response := DBMS_CLOUD.get_response_text(resp);
    SELECT LISTAGG(text, ', ') WITHIN GROUP (ORDER BY ROWNUM)
    INTO v_textfromai
    FROM JSON_TABLE(json_response, '$.imageText.words[*]'
        COLUMNS (
            text VARCHAR2(100) PATH '$.text'
        )
    );
    INSERT INTO aivision_results (id, date_loaded, label, textfromai, jsondata)
    VALUES (SYS_GUID(), SYSTIMESTAMP, p_label, v_textfromai, json_response);
    RETURN v_textfromai;
EXCEPTION
    WHEN OTHERS THEN
        RAISE;
END VISIONAI_TEXTDETECTION;
/ We can also expose the function as a Rest endpoint programmatically using the following: BEGIN
    ORDS.ENABLE_OBJECT(
        P_ENABLED      => TRUE,
        P_SCHEMA      => 'AIUSER',
        P_OBJECT      =>  'VISIONAI_OBJECTDETECTION',
        P_OBJECT_TYPE      => 'FUNCTION',
        P_OBJECT_ALIAS      => 'VISIONAI_OBJECTDETECTION',
        P_AUTO_REST_AUTH      => FALSE
    );
    COMMIT;
END;
/ Analysis and Text Searching of AI Results This architecture also makes analysis and text searching of all AI results convenient and efficient. From here, more processing and analytics can take place. Let’s take a look at three statements that will provide us with an easy-to-use text search of our AI results. First, we create an index for text searches on our table. aivision_results Then, we create a function that searches for a given string using the powerful functionality, or we could additionally/optionally use the package to search multiple tables and return the refcursor of results. contains DBMS_SEARCH Finally, we expose the function as a Rest endpoint. It's that simple. create index aivisionresultsindex on aivision_results(textfromai) indextype is ctxsys.context;
/ CREATE OR REPLACE FUNCTION VISIONAI_RESULTS_TEXT_SEARCH(p_sql IN VARCHAR2) RETURN SYS_REFCURSOR AS refcursor SYS_REFCURSOR;
BEGIN
    OPEN refcursor FOR
        select textfromai from AIVISION_RESULTS where contains ( textfromai, p_sql ) > 0;
    RETURN refcursor;
END VISIONAI_RESULTS_TEXT_SEARCH;
/ BEGIN
    ORDS.ENABLE_OBJECT(
        P_ENABLED      => TRUE,
        P_SCHEMA      => 'AIUSER',
        P_OBJECT      =>  'VISIONAI_RESULTS_TEXT_SEARCH',
        P_OBJECT_TYPE      => 'FUNCTION',
        P_OBJECT_ALIAS      => 'VISIONAI_RESULTS_TEXT_SEARCH',
        P_AUTO_REST_AUTH      => FALSE
    );
    COMMIT;
END;
/ In Conclusion… This was a quick blog showing an architectural pattern for developing data-driven AI Apps by making calls to AI services directly from the database. Thanks so much for reading, and please let me know of any questions or feedback you may have. Also published here

How to Develop Data-Driven AI Apps: A Guide to Making AI Services Directly From the Database

কীভাবে ডেটা-চালিত এআই অ্যাপস ডেভেলপ করবেন: ডাটাবেস থেকে সরাসরি এআই পরিষেবা তৈরি করার জন্য একটি নির্দেশিকা

So entwickeln Sie datengesteuerte KI-Apps: Ein Leitfaden zur Erstellung von KI-Diensten direkt aus der Datenbank

Cómo desarrollar aplicaciones de IA basadas en datos: una guía para crear servicios de IA directamente desde la base de datos

Comment développer des applications d'IA basées sur les données : un guide pour créer des services d'IA directement à partir de la base de données

डेटा-संचालित एआई ऐप्स कैसे विकसित करें: डेटाबेस से सीधे एआई सेवाएं बनाने के लिए एक गाइड

データ駆動型 AI アプリを開発する方法: データベースから AI サービスを直接作成するためのガイド

데이터 기반 AI 앱 개발 방법: 데이터베이스에서 직접 AI 서비스를 만드는 가이드

Como desenvolver aplicativos de IA baseados em dados: um guia para criar serviços de IA diretamente do banco de dados

Как разрабатывать приложения ИИ, управляемые данными: руководство по созданию услуг ИИ непосредственно из базы данных

Veriye Dayalı Yapay Zeka Uygulamaları Nasıl Geliştirilir: Yapay Zeka Hizmetlerini Doğrudan Veritabanından Yapma Kılavuzu

Cách phát triển ứng dụng AI dựa trên dữ liệu: Hướng dẫn tạo dịch vụ AI trực tiếp từ cơ sở dữ liệu

如何开发数据驱动的人工智能应用：直接从数据库制作人工智能服务的指南

Uniswap has revolutionized crypto by introducing the concept of DEXes. However, the platform has also created a new risk for users that provide liquidity to them called Impermanent Losses. Impermanent Losses (IL) are commonly known but do not get serious attention from most liquidity providers. Around 49.5% of the Uniswap v3 liquidity providers not only cannot make a profit due to high IL, but they also lose money at the end. For all of these investors, it will be without investing and wait for a moment when they will grow. This solution looks counter-intuitive when we can see APY around 30% in some Uniswap v3 pools. But the data show that higher APY is also commonly paired with higher ILs. more profitable to hold their tokens In most pools, IL is bigger than the rewards that this pool can generate from fees. On average, IL is 130% of the fees that pools earn. It means that even in pools with high trading volume and big TVL, an average liquidity provider (LP) cannot expect to make any profit. How to be in plus Even if half of the investors are losing money, another half is making them. There are main solutions that investors can use to make their investment profitable: “ ” is known as a popular solution to reduce LPs. Theoretical traders or smart contracts that adjust their position 24/7 could make their investment profitable in most of the pools, but in reality, it is not so straightforward. Data shows that, on average, users, who actively monitor and adjust their positions daily, lose 70% more than those who hold their position for a month or more. Rebalancing your position for months looks more profitable on average. But still, such LPs experience impermanent losses of around 110% of the fees they earn. It means that if the trader is patient enough, he can take a good spot to capture his impermanent losses at the best time when they will be lower than fees. Holding The most promising solution is to to invest in. But choosing such pools is very hard. Except for the size, these pools do not have any pattern. Around 17% of analyzed pools have produced more fees than IL to their holders. They all have mid-size TVLs between $2m — $8m. Pools have purely altcoin pairs on par with stablecoin-to-altcoin pairs. Fees also vary, between 1% and 0.3%. Only advanced analytics can give signs, which pools in reality make money. choose the right pool A consistently profitable solution is also known but is even harder to implement than the previous one. allows for earning positive returns in all pools. It is done by providing liquidity for a single block, absorbing fees from upcoming trades, then instantly removing their position. LPs that use JIT can earn rewards while not experiencing any Impermanent Losses. But such a solution can be profitable only for a large amount of money. Just-In-Time (JIT) liquidity provision I will describe each solution in depth further. Each has pros and cons, which you must consider before investing. Research excludes stablecoin-only pools, as they commonly do not have noticeable impermanent losses. Before we begin This post is based on one of the most in DeFi. Our team analyzed this research to more deeply understand impermanent losses. I decided to share our findings in the hope that they will help someone save money on investments. This post is not financial advice of any kind, and any mentioned protocols and data are provided only for information purposes. Also, while the research was conducted 4.5 months after the Uniswap v3 launch, TVL and volumes in the protocol have grown since then. On the other hand, ratios are very similar and can still be used to understand how to deal with ILs. I extrapolated most of the numbers to make them easier to compare. detailed research on Impermanent Losses Why “active” investors lose more Around a quarter of liquidity providers use rebalancing strategies to reduce impermanent losses, but on average, they only increase their impermanent losses. the a Notice “flash” bar? It is Just-In-Time (JIT) liquidity position. I will explain them in detail further in the post. In the chart, 1d-1w means the account had positions that were active for over one day and less than one week. If the position was adjusted daily or weekly, impermanent losses still significantly outperform fees. All of them, on average, experience IL of around 1.8x higher than the fees they captured. The hourly or monthly adjusted positions look better, but they are still negative. Only positions that have existed for more than a month are close to being positive. They have a ratio of IL to fees of around 1,1x. It gives enough confidence to expect in the long run, holding position can be positive. Investors can find the right moment when the position ratio will return close enough to the initial ratio while collected fees will overcome resulting ILs. But it also means that we cannot expect high returns on investments, with such a high chance of investments being negative. Expected results for investors that actively rebalance their position are much smaller IL than for passive holders, but in reality, we can see the opposite picture. There is always a chance that “active” investors weren’t trying to optimize their position. Still, the most likely variant is that they failed to react to pool condition changes fastly enough and optimize their position precisely enough. Which type of Impermanent Losses are possible to minimize Rebalancing actually cannot remove all ILs, but only a specific type. Types of Impermanent Losses The minimal IL — is a loss that the position suffers while it is in the position price range. While position earns fees, the price in a pool changes, which means asset ratios also change. When the price falls out of the range, the position is only left with one asset and not earning trading fees until the price reenters the range. The difference between the dollar value of the current asset amount and the initial position is a minimal impermanent loss. This IL is unavoidable but can be reduced by decreasing the position price range. The out-of-range IL — is a loss that the position incurred when the pool price exceeds the position price range. While the position ratio is frozen in overweight to the falling asset, the position does not earn any fees. The difference between the dollar value of the initially supplied position and the frozen position is the out-of-range impermanent loss. This loss can be avoided without any losses in profit by withdrawing the position after it goes out of range and repositioning it back with a new ratio. The actual IL — is simply the sum of the minimal IL and the out-of-range IL. It is an end IL that liquidity provider faces. Any rebalancing strategy cannot remove impermanent losses at all, but it can remove out-of-range IL. It means that the strategy captures minimal ILs, which can lead to increasing ILs instead of reducing them in some situations. Also, rebalancing requires a fast reaction to price changes. But even if all actions are implemented properly and in time, it still does not mean that the liquidity provider will have positive returns. In some pools, minimal IL is bigger than the fees this IL produces anyway. The chart shows the difference between IL and fees captured by pools. The blue column shows fees, while the orange shows the actual ILs. A white line marks the minimum IL. You can see already the third by total fees pool, WBTC-WETH, with a 0.3% fee level, has minimal IL bigger than the generated fees. Problems with rebalancing do not stop with choosing the right pool. The first part of rebalancing is withdrawal, which means we “crystalize” the impermanent losses we already have, meaning they are no longer impermanent. At this moment, we already need to have a bigger amount of fees collected than IL, which we capture. Additionally, rebalancing requires the position to be in the ideal size range: After withdrawal, we must swap part of the token to reach the initial position distribution. This swap on big positions can cause substantial price slippage, creating higher losses than IL itself. In the end, we need to deposit a new position, which gives us a total of three transactions per rebalancing cycle. The gas costs of all these transactions can make them useless for small positions. Even if, on each iteration, investors correctly identify the right moment to make rebalancing and always make it with minimal slippage and acceptable transaction fees. It is not enough to reach this theoretical “white line” of minimal impermanent losses. To be able always to sustain these theoretical ILs, the position must be monitored and adjusted 24/7. All these requirements make rebalancing useful only in an automated way. This fact was noticed long ago, and many smart contract protocols already automate rebalancing for their holders. For example, one of the biggest is , specifically focusing on Uniswap v3 rebalancing. https://arrakis.finance Good VS Bad pools You probably already noticed that some pools produced more fees than actual ILs. These pools are rare. Only 3 of 17 analyzed pools have IL lower than fees. It made their LP much more likely to have positive returns. But not only does IL contribute to the probability of having a positive return on investment (ROI) in the pool, but also the volume of transactions and fees level in the pool. It is logical that if the level of the fees we know, IL and volume is hard to predict. In the chart, only 6 pools have a number of positive LPs higher than 50%. On the total average, both negative and positive LPs are between 40–60%. It means if we invest without confidence in pool returns, we can expect a negative return with almost 50% probability. Also, the bigger percentage of positive returns logically causes a bigger percentage of returns to investors. On the left chart, you can see the mean positive return (green) and mean negative return (red) per wallet. As you can see, they’re hard to notice any pattern between pools with higher returns. Positive and negative pools equally have altcoins and stablecoins in pairs. The pool with the best average positive return is FTM-WETH with a 1% fee, which generated a 2.6% ROI. These returns were collected for the first 4.5 months of Uniswap v3 work. This translates to a yield of about 6.9% APY. It is good, but not 20% ROI or 53% APY, which this pool generates directly in fees before IL. You can see a significant difference between the ROI of pools before and after impermanent losses in the next chart. This chart shows the total ROI which pools generated before any impermanent losses. You can see on top WETH-USDT pool with a 0.3% fee. This pool generates over 30% ROI in the first 4.5 months, which is around 80% APY. But also in the previous chart, we sow the average positive return in this pool is 0.9% ROI, which is just 2.4% APY. And it is even good cases. Around half of the users in this pool lost, on average 1.8% of the invested money. It is enough serious sign that before investing in DEXes like Uniswap need check not only average fees APY but also average IL. This problem is known but only partly solved. One of the best yield analytics tools, , currently provides APY and IL analytics only for Uniswap v2. And also, provide APY analytics for Uniswap v3, but it doesn’t cover ILs. https://apy.vision/ https://defillama.com/ This analytics can be valuable to understand and expect APY in other DEXes as well. It, of course, is not enough to make predictions in the future, but enough to decline variants that are lucrative only at first glance. Why does Just-In-Time liquidity not have IL? JIT liquidity is a specific type of liquidity-providing strategy. This strategy is executed only in an automated way, with the help of smart contracts and Flashbot auctions. This strategy requires monitoring all swap transactions in Uniswap v3 pools. When there found a big enough transaction, which had not yet been processed: LP sends a transaction with bigger fees than the target swap transaction. This new transaction deposit liquidity is concentrated exactly around the price range where the swap will be executed. The deposit transaction will be executed before the swap transaction but in the same block. LP lets the swap execute and receives a fee in return for providing liquidity to the swap. After the swap, LP places a withdrawal transaction. This transaction removes all provided liquidity with collected fees. Even if this transaction is executed after the swap, it is still in the same block. There is an example of one of the cycles. All three transactions are executed in the same . block After all the transactions are executed, LP is left with earned fees and a portfolio of tokens, with the ratio changed from the initial one. This difference still introduces the risk of ILs. To remove this risk completely, LP executes alternative swap transactions in parallel on another exchange with a lower swap fee. This swap exchanges tokens in reverse proportion to return the total balance ratio to the initial one. While all of these operations allow removing of impermanent losses, they are also very costly: Two deposit/withdraw transactions with slightly high fees for miners to execute in the target block. Reverse swap fee, together with a small transaction fee. In total, they make JIT liquidity profitable only on high amounts of money. This is probably one of the main reasons why JIT does not get high popularity. The chart shows the total amount of JIT liquidity supplied in Uniswap v3 over time, between 5 May 2021 and 18 July 2022. JIT liquidity provided a total of over $2bn dollars of liquidity, but it is just ~0.3% of all Uniswap v3 liquidity during this period. Even that, over 95% of JIT liquidity was supplied by one single account. While, economically, JIT liquidity can be profitable, there is no available solution to use it. How DeFi is Solving Impermanent Losses Impermanent losses are a pain point for investors in DeFi. However, existing projects can only reduce ILs for liquidity providers in specific scenarios. At Eonian, we believe eliminating impermanent losses is critical to prepare DeFi for mass adoption. New users should not have to worry about IL when investing in DeFi. This is why we decided to build a yield aggregator that distributes liquidity in the most profitable way for liquidity providers. Our protocol will simultaneously use rebalancing, insurance, buffers, and JIT liquidity to remove impermanent losses. More info Research “ ” Impermanent Loss in Uniswap v3 Uniswap team post “ ” Just-in-time Liquidity on the Uniswap Protocol Also published . here

Half of Uniswap v3 Users Lose Money — Here's Why

Information is the foundation of good decision-making. Cryptocurrency innovates at the speed of light. New protocols are constantly coming to market, promising to solve some of mankind's greatest financial troubles. While their intentions might be good, the result is a cluttered, noisy environment that promotes FOMO & makes it nearly impossible to make sound decisions. In order to succeed in the wild, volatile, & often opaque world of cryptocurrency it is necessary to quiet the noise & hone in on real signals. The best way to do that is by building out your own Crypto ToolKit. Quick Notes on building a Crypto ToolKit: -) Do not use every single tool (unless you’re a crypto savant) -) Review each resource/tool individually (5–10 min per resource) -) Select only the ones that appeal most to your specific preferences As you might already be aware, the crypto-verse is highly fragmented in the sense that there are many different chains & protocols. Given that not all chains & protocols are interoperable (yet), all of the tools listed provided below are will be as general & multi-purpose as possible. Without further ado, Let's dive in: 📊 Market Data 📊 The starting point of almost every person’s crypto journey, Market Data providers are listing sites that aggregate all open financial digital asset data. All of the fundamental economic elements, Price, Volume, Circulating supply, Orderbooks, market dominance, historic data, & everything else. 1) Token Terminal - https://tokenterminal.com/ 2) Messari - https://messari.io/ 3) Crypto Slate - https://cryptoslate.com/coins/ 4) CoinGlass - https://www.coinglass.com/ Coin Market Cap -** 5) https://coinmarketcap.com/ - 6) Coin Gecko https://www.coingecko.com/ 7) Coin Metrics - https://coinmetrics.io/ 8) Coinlib - https://coinlib.io/ 9) Crypto Compare - https://www.cryptocompare.com/ 10) CoinCap - https://coincap.io/ - 11) CoinRanking https://coinranking.com/ 12) Coin360 — https://coin360.com/ — 13) Bitgur https://bitgur.com/ — 14) Nomics https://nomics.com/ - 15) CryptoWatch https://cryptowat.ch/ 16) BraveNewCoin — https://bravenewcoin.com/ 17) LiveCoinWatch — https://www.livecoinwatch.com/ 18) MultiCoinCharts - https://www.multicoincharts.com/home/ 19) Nansen - https://www.nansen.ai/ 20) CoinPaprika - https://coinpaprika.com/ 21) CoinAlpha - https://coinalpha.app/ 22) CoinScope - https://www.coinscope.co/ 23) DropsTabs - https://dropstab.com/ 24) CryptoView - https://www.cryptoview.com/ 25) CoinCarp - https://www.coincarp.com/ 26) Mytokencap - https://www.mytokencap.com/ 27) CoinCodex — https://coincodex.com/ — 28) CoinLobster https://coinlobster.com/ 💱 DEFI Dashboard & Aggregators💱 Decentralized finance is an incredibly broad term that includes a wide range of applications that vary based on functionality & intended purpose. While the majority of current DEFI activity happens across 4 blockchains (Ethereum, Binance Smart Chain, Polygon & Solana) there are so many up & coming alternative platforms such as Optimism & Arbitrum that have been garnering attention that for the sake of brevity I will only list the few products that I am personally familiar with. 1) Dapp Radar — https://dappradar.com/defi 2) DEFI Llama — https://defillama.com/ 3) DEFI Prime — https://defiprime.com/ 4) DEFI Pulse — https://www.defipulse.com/ 5) YieldWatch - https://www.yieldwatch.net/ 6) ZapperFI - https://zapper.xyz/ 7) Zerion - https://zerion.io/ 8) InstaDapp - https://instadapp.io/ 9) Frontier - https://www.frontier.xyz/ 10) Apebored - https://apeboard.finance/ 11) Debank - https://debank.com/ - 12) APYvision https://app.apy.vision/ - 13) Plasma Finance https://plasma.finance/ - 14) Farm Army https://farm.army/ - 15) DEFISaver https://defisaver.com/ - 16) Gnosis Safe https://gnosis-safe.io/ - 17) DefiYield https://defiyield.app/ 18) Coin98 - https://coin98.com/ 19) AstroTools - https://astrotools.io/ 20) MultiFarm - https://www.multifarm.fi/ 21) OctoFi - https://octo.fi/ 22) Parallel - https://parallel.fi/ 23) Exponential - https://exponential.fi/ 24) WalletDefi - https://walletdefi.app/ 25) DEFImarketCap - https://defimarketcap.io/ 26) QuantifyCrypto - https://quantifycrypto.com/ 📰 News 📰 Crypto is becoming all the more politicized as time goes on. An absurd amount of crypto/DEFI news sites pop up constantly & push their own agendas. The best way to source news (in any industry) is by having a few reliable individual outlets & cross-referencing information across them all. 1) CoinDesk - https://www.coindesk.com/ 2) Coin Telegraph - https://cointelegraph.com/ 3) Bitcoin Magazine - https://bitcoinmagazine.com/ 4) The Cryptonomist - https://en.cryptonomist.ch/ 5) Be In Crypto - https://beincrypto.com/ 6) CoinGape - https://coingape.com/ 7) CryptoSlate - https://cryptoslate.com/ 8) Crypto Briefing - https://cryptobriefing.com/ 9) CryptoDaily - https://cryptodaily.co.uk/ 10) Crypto Potato - https://cryptopotato.com/ 11) BraveNewCoin - https://bravenewcoin.com/ 12) Coin Journal - https://coinjournal.net/ 13) Coin Speaker - https://www.coinspeaker.com/ 14) CoinCu - https://news.coincu.com/ 15) Bitcoinist - https://bitcoinist.com/ 16) CryptoPanic - https://cryptopanic.com/ 17) Crypto News - https://cryptonews.com/ 18) U Today - https://u.today/ 19) CoinGraph - https://www.coingraph.news/ - 20) Decrypt https://decrypt.co/ - 21) Blockworks https://blockworks.co/ - 22) CryptoCapitalNews https://cryptocapitalnews.com/ — 23) Protos https://protos.com/ 👛 Wallets 👛 Wallets are the foundation for getting involved & staying in crypto. Wallets are exactly what they sound like, places to securely store/view your crypto asset holdings. Wallets typically come in 2 flavors, Chain-specific & Chain-agnostic. Chain-specific wallets are tied to a single blockchain platform (i.e ETH/BNB). These will only be able to store their native chain tokens, but allow the users to interact with on-chain assets/applications. Whereas chain-agnostic wallets might allow their users to store crypto assets from any blockchain, but that is the limit of their capabilities as they are usually single-purpose only (storage) 1) MetaMask - https://metamask.io/ 2) MEW - https://www.myetherwallet.com/ 3) Dynamic - https://www.dynamic.xyz/ 4) XDEFI - https://www.xdefi.io/ 5) Enkryptic - https://www.enkrypt.com/ 6) Argent - https://www.argent.xyz/ 7) Dopamine - https://dopamineapp.com/ 8) TrustWallet - https://trustwallet.com/ 9) Eidoo - https://eidoo.app/ 10) Dapppocket - https://www.dapppocket.io/ 11) Atomic Wallet - https://atomicwallet.io/ 12) Bread Wallet - https://brd.com/ 13) Guarda - https://guarda.com/ 14) Exodus - https://www.exodus.com/ 15) ZenGo - https://zengo.com/ 16) Coinomi - https://www.coinomi.com/en/ 17) Luno - https://www.luno.com/en-US 18) Ambire - https://www.ambire.com/ 19) Jaxx - https://jaxx.io/ 20) Edge - https://edge.app/ 21) Freewallet - https://freewallet.org/ 22) Coinbase Wallet - https://www.coinbase.com/wallet 23) Coin Wallet - https://coin.space/ 24) Ballet Wallet - https://www.ballet.com/coins/ 25) Unstoppable - https://unstoppable.money/ 26) Brave Wallet - https://brave.com/wallet/ 27) BitPay - https://bitpay.com/wallet/ 28) Crypto.com - https://crypto.com/us/defi-wallet 🔎 Explorers 🔍 Explorers are software tools that allow users to inspect the activity that has happened/is happening on-chain. Explorers are typically built-to-blockchain, meaning every chain needs its own explorer. Given the girth of data these tools access, explorers tend to be multi-purpose tools (i.e. data providers/ analytics dashboards/compliance). 1) Ethscan - https://etherscan.io/ 2) Bitquery - https://explorer.bitquery.io/ 3) Blockchair - https://blockchair.com/ 4) Tokenview - https://tokenview.io/ 5) Blockcypher - https://www.blockcypher.com/ 6) GetBlock - https://getblock.io/explorers/ 7) ViewBlock - https://viewblock.io/ - 8) BreadCrumbs https://www.breadcrumbs.app/ 9) Beaconchain - https://beaconcha.in/ 10) Blockstream - https://blockstream.info/ 11) Ethplorer - https://ethplorer.io/ 12) OXT - https://oxt.me/ 13) Xscan - https://xscan.io/ 14) Sochain - https://sochain.com/ 15) OkLink - https://www.oklink.com/en 16) Pingpub - https://ping.pub/ ⛓ On-Chain Data ⛓ On-chain data tools are closely related to explorers in the sense that they allow users to inspect the activity that has happened / is happening on-chain. The main difference between the On-chain Data tools & Explorers is the UI & granularity of data they provide.*The main difference between On-Chain data providers & traditional market data is the absence of information that is external to any blockchain. 1) Glassnode - https://glassnode.com/ 2) Dune Analytics - https://dune.com/home 3) IntoTheBlock - https://www.intotheblock.com/ - 4) Mempool https://mempool.space/ 5) Artemis - https://www.artemis.xyz/ - 6) Parsiq https://www.parsiq.net/ 🌉 Bridges 🌉 Within the context of cryptocurrency, a bridge is a software tool or protocol that allows otherwise non-compatible blockchains a method to transfer digital assets between one another. 1) MultiChain - https://multichain.org/ 2) CrossChain Bridge - https://www.crosschainbridge.org/ 3) Celer C Bridge - https://cbridge.celer.network/1/56/USDC 4) Synapse - https://synapseprotocol.com/ 5) Connext - https://bridge.connext.network/USDC-from-ethereum-to-binance 6) Relay - https://www.relaychain.com/ 7) Symbiosis - https://symbiosis.finance/ 8) Umbria - https://umbria.network/ 9) EverBridge - https://www.everrise.com/everbridge/ 10) Bungee - https://www.bungee.exchange/ 11) Portal Bridge - https://www.portalbridge.com/#/transfer 12) Nomad - https://www.nomad.xyz/ 13) Hop - https://hop.exchange/ 14) PolyBridge - https://bridge.poly.network/ 15) SolarBeam - https://app.solarbeam.io/bridge 16) Orbitchain - https://orbitchain.io/about 💹 DEXes 💹 Decentralized Exchanges are the new financial trading paradigm. Generally free from traditional regulatory scrutiny, DEXs are places to buy, sell, swap & provide liquidity for crypto assets. 1) UniSwap - https://uniswap.org/ 2) 1inch - https://app.1inch.io/ 3) DexTools - https://www.dextools.io/app/en 4) DexGuru - https://dex.guru/ 5) GeckoTerminal - https://www.geckoterminal.com/ 6) PancakeSwap - https://pancakeswap.finance/ 7) TraderJoe - https://traderjoexyz.com/avalanche 8) Arken - https://swap.arken.finance/ 9) Flooz - https://flooz.trade/trade 10) Balancer - https://balancer.fi/ 11) Bancor - https://home.bancor.network/ 12) Airswap - https://www.airswap.io/#/ - 13) Sushiswap https://www.sushi.com/ 14) Paraswap - https://www.paraswap.io/ 15) Jupiter - https://jup.ag/ 16) ThorChain - https://thorchain.org/ 17) WooFi - https://fi.woo.org/ 18) DYDX - https://trade.dydx.exchange/ - 19) Slingshot https://app.slingshot.finance/swap/ETH 20) Apex Exchange - https://www.apex.exchange/ 21) Osmosis - https://app.osmosis.zone/ 22) Curve - https://curve.fi/#/ethereum/swap 23) Tokenlon - https://tokenlon.im/instant 24) Swapz — https://www.swapz.app/ 📅 Calendars & Event Schedules 📅 Keeping up with the Kardashians is much easier than keeping up with your favorite crypto protocol; the question then becomes HOW do you plan to manage the onslaught on events? While there is no easy answer, using a Crypto Calander will definitely help: 1) CoinMarketCal - https://coinmarketcal.com/en/ 2) CoinDar - https://coindar.org/ 3) CoinsCalendar - https://coinscalendar.com/ 4) CryptoTotem - https://cryptototem.com/ico-calendar/ 5) CoinCarp events - https://www.coincarp.com/events/ 6) VestLabs - https://vestlab.io/calendar 🪂 Airdrops 🪂 The most popular tool for distribution in the entire industry, airdrops are incentives that are offered/provided at the earlier stages of a project’s development to help bootstrap it’s alpha community members. 1) Earnifi - https://earni.fi/ 2) airdropsio - https://airdrops.io/ 3) AirDropAlert - https://airdropalert.com/ 4) AirdropsLive - https://airdrops.live/ 5) AirdropBob - https://www.airdropbob.com/ 6) AirDropAddict - https://airdropaddict.com/ 7) CoinAirDrops - https://coinairdrops.com/ 8) ICOdrops - https://icodrops.com/ 💸 Grants 💸 Grants are a necessary component to helping startups & innovators get their ideas off the ground. In the world of crypto & blockchain grants are usually offered & deployed on a chain/case-specific basis. In order to be properly up to date on grant opportunities it is best to be actively involved in individual projects and/or blockchains. 1) Blockchain Grants - https://www.blockchaingrants.org/ 2) GitCoin - https://gitcoin.co/grants/ 3) Consensys - https://consensys.net/grants/ — 4) Concoridum https://concordium.com/grants/ 5) Web3 grants - https://web3grants.net/ 6) Web3Foundation - https://web3.foundation/grants/ 7) Ramp Rate - https://ramprate.com/blockchain-grant-opportunities/ - 8) Unigrants https://www.unigrants.org/ 9) Aave grants - https://aavegrants.org/ 10) Near grants - https://near.org/ecosystem/get-funding/ 11) Eth Foundation - https://esp.ethereum.foundation/ 12) Zcash grants - https://grants.zfnd.org/ 13) Alchemy Dev grants - https://www.alchemy.com/developer-grant-program — 14) dydx grants https://www.dydxgrants.com/ 15) Solana grants — https://solana.org/grants 16) Gate grants - https://www.gate.io/grants 17) Cryptoneur - https://www.cryptoneur.xyz/en/web3-grants 🛡 Audits 🛡 Audits are used to assess the quality/security of a protocol. When considering a protocol to use, it is nearly impossible to understand all of the associated risk factors, even for professionals. It is crucial that every serious project employs some form of audit for every individual piece of software. 1) Certik - https://www.certik.com/ 2) Consensys Dilligence - https://consensys.net/diligence/ 3) Hacken - https://hacken.io/services/smart-contract-security-audit/ 4) Open Zepplin - https://www.openzeppelin.com/security-audits 5) Trail of Bits - https://www.trailofbits.com/ 6) Slowmist - https://www.slowmist.com/ 7) Solidified - https://solidified.io/ 8) Chain Security - https://chainsecurity.com/smart-contract-audit-reports/ 9) PeckShield - https://peckshield.com/#home 10) Halborn - https://halborn.com/ 11) Quantstamp - https://quantstamp.com/ 12) Solidity Finance - https://solidity.finance/ - 13) Immune Bytes https://www.immunebytes.com/ 14) NumenCyber - https://www.numencyber.com/ 🌊😎 Cool & Honorable Mentions 😎🌊 There are always exotic instruments & services being built to help make sense of the wild world of crypto; some of these are difficult to classify but still deserve attention for serving their own unique & useful purpose(s). 1) Fiat Leak - https://fiatleak.com/ 2) Everest - https://everest.link/ 3) Revoke.cash - https://revoke.cash/ 4) CryptoQuant - https://cryptoquant.com/asset/btc/summary - 5) CryptoFees https://cryptofees.info/ - 6) Hashrate Index https://data.hashrateindex.com/asic-index-data — 7) LunarCrush https://lunarcrush.com/home — 8) Santiment https://santiment.net/ 9) Gas Fee Calculator - https://www.cryptoneur.xyz/en/gas-fees-calculator This list is by no means exhaustive, there are thousands of other tools/resources out there. Odds are, you do not even need 1/10 of those on this list; so my recommendation is don’t waste time wishing for better tools — learn how to masterfully use a few key ones provided above. If you went through the entire list, you might have noticed that there are no CEFI (centralized financial) tools on this list. This was done on purpose, given the madness that transpired with centralized exchanges/service providers throughout 2022. If you would like me to add any specific resource that might be beneficial to others, just drop a comment below or send me a DM on Twitter. Otherwise; I hope you find these tools helpful in building your knowledge & financial freedom. Thank you so much for reading! 🥂🚀 See you in the next Bull Market 🚀🥂 Also published here.

200+ Essential Crypto & DEFI Tools for 2023

Bear Markets is hard. Many individuals lost their cash made during the bull market. This time is likewise hard for our psychological wellness. This is particularly valid for individuals who are new to crypto. This is a cyclical market. We have ups and downs. Wealth creation and destruction. Hope and sorrow. The bear market Figuring out how this market functions can give you the edge. Regardless of whether you miss this bull market or lost all your gains in this bear market, you can get ready for the following cycle. To do this, begin with the past. Dissect your errors and gain from them. Try not to stop here. Gain from others' mix-ups. In this article, you can find learnings from that bear market. I broke down the most popular investors’ errors and made from them mental models, which you can apply to your investment strategy. You don't have to be always right Most investors want to be always right. This is not possible in any environment. The market is not fully predictable. No investor or trader has only good investments.  There is nothing wrong with that. You don’t have to be always right. You need to be better than average market participants. You can have 9 bad bets and 1 which paid off for all of them. Shift your focus from position management to portfolio management. Choose your evaluation period. Don’t check your portfolio returns every hour. Short-term variability is not manageable. Focus on the term period. You are not diversified When there is anxiety in the market, diversification is most important. It may be too late to do this when you are anxious about the state of the market. . Also, estimate how much time will be needed to manage these positions. Adapt the degree of diversity to your risk tolerance and appetite Stablecoins, digital gold, and synthetic assets come to mind. Also, take into account various blockchains for holding these tokens. Diversification is not holding a bunch of different cryptocurrencies. The majority of them are correlated. You should own a variety of digital assets in addition to crypto tokens. There is an offline world We don't actually reside in the metaverse. During a bull market apex, Apple was worth more than the whole cryptocurrency market. We still have time. Even if you believe in cryptocurrency, don't put all of your money in it. Boost your portfolio by including bonds, gold, equities, or real estate. Learning the fundamentals of conventional finance is also a smart idea during a bad market. You may still be a passive investor even if you are not an expert in stocks. Start with diverse industry and regional . When you discover your market advantage, switch to . ETFs active investing Your altcoins never reach ATH again Most of the time, speculation drives token pricing. Most likely, your tokens also won't make a difference in the real world. Most initiatives are motivated by marketing rather than actual use. Most of your projects don't have any use cases. Recognize the factors that contributed to the demand for your tokens and the potential future buyers. The basis of the crypto investment is tokenomics. This frequently distinguishes between wise and foolish investments. Snapshots of the top projects for each week are available on . Look at how rapidly the story changes. The price is almost always declining and many projects are now forgotten. Why it will be different with your tokens? Coinmarketcap The market moves in cycles We are aware of it, but we are unprepared. drives the majority of decisions. Since the price is increasing, will it continue to do so in the future? Don't extrapolate from previous weeks. Your charts can fool you. Observe market trends and the larger picture. Momentum You may profit from the upcoming bull market with the aid of these instructions. Every week, determine the worth of your portfolio and ask yourself, "How much can I lose?" and “Am I prepared for a bear market?”. This is time to learn Bull markets are insane. Every day, new projects are launched. New farming prospects are also being presented to us. is incredibly simple to experience. Bear markets are sluggish. With , projects would rather wait until better times. FOMO ICOs Additionally, are far less volatile and lower. A moment for learning is now. Even if the next bull market never materializes, you will always remember this information. DeFi yields Learn new meta-skills, revise your plan, and identify emerging trends. You don't understand what you are buying Each token has a project attached to it. The majority of individuals are unaware of their investment. If there isn't a token for this project, will you still invest directly in it? Only in a bull market is this effective. Consider this to be your project (you are the CEO). How will this alter your viewpoint? What are the project's advantages and disadvantages? What about chances and dangers? Better comprehension leads to better financial choices. https://twitter.com/CryptosEngineer/status/1521558062686883845 Most projects work only during the bull market The following conditions are necessary for DeFi initiatives to succeed: Most cryptocurrency projects use DeFi. Even the majority of games are simply gamified extensions of the payment system. an inflow of new cash to crypto, many tokens swap. The whole DeFi technology is powered by these two operations. In a bull market, this is profitable. Profits are substantially less during a bad market. Check out this , as an illustration, to compare the revenue during several market cycles. project Stablecoins are not equal They carry greater risk than money. You want to maintain a diverse portfolio of in various: Never presume that any stablecoin is secure. stablecoins Tokens Protocols Chains Strategies is a prime illustration of this. We also saw many hacks. Your whole earnings might be wiped out by a lack of diversity. UST bridge Understand your mistakes Everyone is making mistakes. The distinction is that it may be applied by those who are successful. The following categories apply to all errors: (too little information, incorrect information, wrong analytical process) Analytical/Cognitive (greed and fear, ego and envy, willingness to suspend disbelief and extrapolate past) Emotional/Psychological Use this to investigate the root reasons for your prior errors. Create an investing diary if you don't already have one. To become a better investor, you will be frequently evaluating it. Project Value = Story + Numbers Every metric we employ in a valuation ( , Tx number, user growth, etc.) must be based on a narrative about the project. Every detail we mention about a project, such as its incredible staff, outstanding platform, or wonderful community, must be represented by a number. TVL Most individuals only pay attention to one element of this equation. Some people are simply posting figures out of context from other sources. Others invent stories about how this undertaking will alter the course of history. You cannot become a successful investor with either choice. Fundamental vs technical analysis The relative merits of fundamental and technical analysis are always under discussion. If you know how to utilize them, both may be effective. Select your team. The short-term fluctuation of value could not be explained by the basics of fundamental analysis. Are you a trader or an investor? can be effective because market players frequently base their judgments on emotions and because prices represent all available information. , on the other hand, works in competitive marketplaces. The general agreement is that yes, but over a longer period of time. Technical analysis Fundamental analysis https://twitter.com/CryptosEngineer/status/1521142480837627912 You don't have an investment strategy How would you choose these pricing ranges? You need a different approach to do this. A strategy is a framework you can use to get where you want to go. Establishing and managing your portfolio is one of these objectives. Buying low and selling high (contrarian) or buying high and selling even higher (momentum) is not a strategy. Consider carefully why you have been involved in the cryptocurrency industry as well as your goals and desired outcomes. You are especially guided by strategy during these unsettling times. https://twitter.com/CryptosEngineer/status/1520776920148983813 You are not managing your portfolio can be simplified using dedicated tools. If you have too many positions you won't be able to manage them, especially during bear markets. Portfolio management Check: https://app.zerion.io/ https://debank.com/ https://zapper.fi/ https://apeboard.finance/ https://app.apy.vision/ I also use Excel to track all my investment (not just crypto) in one place. https://twitter.com/CryptosEngineer/status/1523323205192015873 You are not managing your risk Most investors think that they have a higher risk appetite and capacity. - The total amount of risk that you can tolerate Risk capacity - The degree to which you will embrace negative risk in pursuit of your objectives Risk appetite If you look at how you responded to changes in your portfolio, you can quantify your risk tolerance. The investing diary is again your friend in this situation. Check also . Risk Management 101 https://twitter.com/CryptosEngineer/status/1522641852301684738 What Next? In future posts, I will discuss the new process . Cryptocurrency Investment Framework Stay tuned and follow me on to get a sneak peek of what is coming! Twitter

Bear Market Mental Models - Are You Ready for the Next Bull Run?

Modelos mentales del mercado bajista: ¿está listo para la próxima carrera alcista?

Modèles mentaux du marché baissier - Êtes-vous prêt pour la prochaine course haussière ?

भालू बाजार मानसिक मॉडल - क्या आप अगले बुल रन के लिए तैयार हैं?

弱気市場のメンタル モデル - 次の強気相場の準備はできていますか?

Modelos mentais do mercado de urso - você está pronto para a próxima corrida de touros?

Các mô hình tinh thần của thị trường gấu - Bạn đã sẵn sàng cho đợt tăng giá tiếp theo chưa?

熊市心理模型——你准备好迎接下一轮牛市了吗？

Israeli startups embrace AI across multiple industries with great success – meanwhile investment in these companies exceeded $25 billion in 2021 alone. In recent years, Israel has established itself as a leading tech nation, raising capital and growing more unicorns than any other country (per 100,000 citizens). As a matter of fact, it’s precisely Israel that’s ahead of most other countries, both in terms of the number of successful AI startups, as well as the overall investment these businesses have been able to attract. All that despite the country’s small size by comparison. 2014-2019 The numbers speak for themselves. According to , Israel had around 500 registered AI companies that managed to raise over $500 million in funding in 2014. Over the next 4 years, these figures grew dramatically to the point that Israel counted as many as 1150 AI businesses in 2018. Together these companies raised an impressive $2.25 billion in funding. If we also include Israeli high-tech and deep-tech companies, the investment figures climb to a whopping $6 billion. Quite impressive for a nation of 9 million citizens that’s roughly the size of New Jersey. The Times of Israel In 2018, among the AI go-getters from Israel were those that worked on cybersecurity (Israel’s forte traditionally: around 20% of total funding,), healthcare (around 15% of total funding), followed by law, computer vision, and autonomous vehicles among other areas. What’s more, according to , of the roughly 6,800 tech startups in Israel that year, almost a quarter were involved in AI in one form or another – a baffling figure compared to the much humbler stats coming from other nations. Start-Up Nation Finder 2020-today Today, Israel is showing no signs of stopping or even slowing down – at least so far – regardless of the Pandemic. Israel continues to thrive. According to the , Israeli tech startups raised $10,3 billion in 2020, followed by $25,6 billion through 773 investment rounds last year. The latter figure is a 400% increase from 3 years ago and a 150% increase from 2020, mainly due to “the large number of deals over $100 million.” IVC-Meitar report Predictably, cybersecurity topped the list again with $6.5 billion in investment for all 4 quarters. The last quarter of 2021 was the strongest, having raised $8 billion across multiple industries. It’s important to note that 65% of that funding is foreign, with the US dominating both seed-stage and series A rounds. This comes as no surprise knowing the entrepreneurial spirit of the American businessman, combined with the fact that post-money valuation of the Israeli startups in 2021 rose a staggering 650% from the previous year alone. According to Mike Rimon from Meitar Law Offices, “2021 was the year of the Big Bang” for Israeli IPOs in the US. In addition to that, 13 Israeli companies went for the SPAC option (Special Purpose Acquisition Companies), which is another way for a private company to go public on the NYSE or Nasdaq. They successfully raised just under $3 billion in total. Israeli tech startups up close Among noteworthy tech startups from Israel today are , a B2B payment platform, and , a workforce management platform, that managed to raise $250 million each in their series D rounds in the fall of last year. Both companies are now valued at around $4 billion each. Melio Papaya Global , an Israeli cybersecurity company, managed to raise a record $400 million in funding last December. This is considered to be the biggest investment deal to date within the global cybersecurity sector. SoftBank led the round, a Tokyo-based conglomerate that focuses on investing in emerging technologies. SoftBank also gave $200 million in funding to , an Israeli IoT company backed by Amazon Web Services that utilizes Machine Learning (ML) for cloud intelligence. Claroty Wiliot Another 3 AI startups from Israel received over $350 million each in funding through several investment rounds. One of them is (formerly AnyVision), a platform with 3 flagship products that provides Fortune 500 brands with face-recognition and video-surveillance technology ($352 million). is another example, an AI company that helps businesses with audio, video, and text transcription and captioning ($550 million). And there’s , a conversation intelligence platform for B2B sales that uses ML and Natural Language Processing (NLP) to train and assist sales teams ($583 million). They’re also active in to ML training data. Oosto Verbit Gong.io bringing publicity CES top picks A few AI startups from Israel also made an appearance at the highly regarded (CES) in Las Vegas earlier this year – BioBeat, OrCam, and Cecilia among others. , a health-AI company, is focused on generating clinical data. One of its products that’s popular among doctors in Israel – and now spreading overseas – is an AI-powered patch that takes the patient’s measurements. From blood pressure to oxidation rate, there are 14 vital signs in total. Consumer Electronics Show BioBeat is an extraordinary company that helps visually impaired and legally blind individuals to walk in public and identify various places and items. One of the company’s groundbreaking products is MyEye Pro, an AI-powered navigation device that transforms text into speech for the wearer and whispers to them when a familiar face is recognized. OrCam On a less serious note, is the first AI-driven interactive bartender named after a speakeasy legend, Mary Louise Cecilia Guinan. The eight-foot-tall virtual mixologist uses cutting-edge voice-recognition technology to interact with the customer and offer them suggestions based on their needs, or even their mood. The robolady says: “I’m chattier than most robots you’ll meet. I talk, I listen, and I’m goddamn hilarious.” She can serve up to 120 cocktails per hour. Cecilia Takeaway As we’ve seen, Israel is setting a fine example for other countries to follow. For the past decade, Israeli startups have raised tens of billions of dollars – some of it for the tech sphere in general and some for the AI field in particular. Last year alone, the figures exceeded $25B. This year the numbers promise to reach a new peak, thanks to geo-political changes, Israel promises to be fertile ground for successful products. It will be interesting to see whether new Israeli tech startups will continue to emerge over the next few years. Given the current pace, what’s even more interesting is how the current high-innovation trend will evolve – both in terms of the AI products offered to the public and the amount of capital invested in these solutions.

Israel is Swiftly Becoming the World’s Leading AI Hub as Investments Continue to Surge in 2022

One school using the software saw that a student’s face was captured more than 1,000 times during the week By Alfred Ng In 2019, the Santa Fe Independent School District in Texas ran a weeklong pilot program with the facial recognition firm AnyVision in its school hallways. With more than 5,000 student photos uploaded for the test run, AnyVision called the results “impressive” and expressed excitement at the results to school administrators. “Overall, we had over 164,000 detections the last 7 days running the pilot. We were able to detect students on multiple cameras and even detected one student 1100 times!” Taylor May, then a regional sales manager for AnyVision, said in an email to the school’s administrators. The number gives a rare glimpse into how often people can be identified through facial recognition, as the technology finds its way into , stores, and public spaces like sports arenas and casinos. more schools We were able to detect students on multiple cameras and even detected one student 1100 times! -An AnyVision regional sales manager May’s email was among hundreds of public records reviewed by The Markup of exchanges between the school district and AnyVision, a fast-growing facial recognition firm based in Israel that , including schools, hospitals, casinos, sports stadiums, banks, and retail stores. One of those retail stores is Macy’s, which , according to Reuters. Facial recognition, purportedly AnyVision, is also being used by a supermarket chain in Spain to detect people with prior convictions or restraining orders and prevent them from entering 40 of its stores, according to research published by the . boasts hundreds of customers around the world uses facial recognition to detect known shoplifters European Network of Corporate Observatories Neither Macy’s nor supermarket chain Mercadona responded to requests for comment. The public records The Markup reviewed included a for AnyVision’s software called “Better Tomorrow.” The manual contains details on AnyVision’s tracking capabilities and provides insight on just how people can be identified and followed through its facial recognition. 2019 user guide The growth of facial recognition has raised privacy and civil liberties concerns over the technology’s ability to constantly monitor people and track their movements. In June, the European Data Protection Board and the European Data Protection Supervisor called for a , warning that “deploying remote biometric identification in publicly accessible spaces means the end of anonymity in those places.” facial recognition ban in public spaces Lawmakers, privacy advocates, and civil rights organizations have also pushed against facial recognition because of error rates that disproportionately hurt people of color. A 2018 highlighted how facial recognition technology from companies like Microsoft and IBM is consistently less accurate in identifying people of color and women. research paper from Joy Buolamwini and Timnit Gebru In December 2019, the National Institute of Standards and Technology also found that the majority of . There have been at least . facial recognition algorithms exhibit more false positives against people of color three cases of a wrongful arrest of a Black man based on facial recognition as a watchlist-based facial recognition program, where it only detects people who are a known concern. Stores can buy it to detect suspected shoplifters, while schools can upload sexual predator databases to their watchlists, for example. “Better Tomorrow” is marketed But AnyVision’s user guide shows that its software is logging all faces that appear on camera, not just people of interest. For students, that can mean having their faces captured more than 1,000 times a week. AnyVision’s user guide said it keeps all records of detections for 30 days by default and allows customers to run reverse image searches against that database. That means that you can upload photos of a known person and figure out if they were caught on camera at any time during the last 30 days. The software offers a “Privacy Mode” feature in which it ignores all faces not on a watchlist, while another feature called “GDPR Mode” blurs non-watchlist faces on video playback and downloads. The Santa Fe Independent School District didn’t respond to a request for comment, including on whether it enabled the Privacy Mode feature. “We do not activate these modes by default but we do educate our customers about them,” AnyVision’s chief marketing officer, Dean Nicolls, said in an email. “Their decision to activate or not activate is largely based on their particular use case, industry, geography, and the prevailing privacy regulations.” AnyVision boasted of its grouping feature in a “Use Cases” document for smart cities, stating that it was capable of collecting face images of all individuals who pass by the camera. It also said that this could be used to “track [a] suspect’s route throughout multiple cameras in the city.” The Santa Fe Independent School District’s police department wanted to do just that in October 2019, according to public records. In an email obtained through a public records request, the school district police department’s Sgt. Ruben Espinoza said officers were having trouble identifying a suspected drug dealer who was also a high school student. AnyVision’s May responded, “Let’s upload the screenshots of the students and do a search through our software for any matches for the last week.” The school district originally purchased AnyVision after , with hopes that the technology would prevent another tragedy. By January 2020, the school district had uploaded 2,967 photos of students for AnyVision’s database. a mass shooting in 2018 James Grassmuck, a member of the school district’s board of trustees who supported using facial recognition, said he hasn’t heard any complaints about privacy or misidentifications since it’s been installed. You’ve built a system that’s essentially like putting an ankle monitor on all your kids. Clare Garvie, Georgetown University Law’s Center on Privacy and Technology “They’re not using the information to go through and invade people’s privacy on a daily basis,” Grassmuck said. “It’s another layer in our security, and after what we’ve been through, we’ll take every layer of security we can get.” The Santa Fe Independent School District’s neighbor, the Texas City Independent School District, also purchased AnyVision as a protective measure against school shootings. It has since been used in attempts to identify a kid who had been licking a neighborhood surveillance camera, to kick out an expelled student from his sister’s graduation, and to ban a woman from showing up on school grounds after an argument with the district’s head of security, . according to WIRED “The mission creep issue is a real concern when you initially build out a system to find that one person who’s been suspended and is incredibly dangerous, and all of a sudden you’ve enrolled all student photos and can track them wherever they go,” Clare Garvie, a senior associate at the Georgetown University Law Center’s Center on Privacy & Technology, said. “You’ve built a system that’s essentially like putting an ankle monitor on all your kids.” Originally published as " “ with the Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license This Manual for a Popular Facial Recognition Tool Shows Just How Much the Software Tracks People

Is There a Valid Cause for Using Facial Recognition Tools in Schools

Liquidity provisioning in Uniswap V3 presents a stochastic optimal control problem with a well-defined utility function to maximize. This article introduces an innovative framework for intelligent liquidity provisioning, utilizing a combination of agent-based modeling and reinforcement learning. Our framework provides a robust and adaptive solution for optimizing liquidity provisioning strategies. The Uniswap V3 model mimics real-world market conditions, while the agent-based model (ABM) creates an environment for simulating agent interactions with Uniswap V3 pools. The reinforcement learning agent, trained using deep deterministic policy gradients (DDPG), learns optimal strategies, showcasing the potential of machine learning in enhancing DeFi participation. This approach aims to improve liquidity providers’ profitability and understanding of CFMM markets. Content Overview Introduction Intelligent liquidity provisioning framework Components of Intelligent Liquidity Provisioning Framework Agent-based simulator Reinforcement Learning Model Limitations Future Work Conclusion Resources References Introduction In my previous article on market making we explored the mechanics and strategies of market making in traditional financial markets. Building upon those insights, this article introduces an innovative framework for intelligent liquidity provisioning in the context of Uniswap V3. As mentioned in our prior research, our goal was to extend our understanding of market dynamics and liquidity management in decentralized finance ( , specifically through the development of the Intelligent Liquidity Provisioning Framework. [Market Making Mechanics and Strategies], DeFi) Decentralized finance (DeFi) has undergone remarkable growth, introducing innovative financial products and services accessible to a global audience. Uniswap V3, at the forefront of this innovation, has revolutionized liquidity provisioning with its concentrated liquidity feature. However, this advancement brings forth complex decision-making challenges for liquidity providers. This article introduces a comprehensive framework designed to address these challenges, offering a simulated environment for studying and optimizing liquidity provisioning strategies. Our framework comprises three key components: the Uniswap V3 model, an agent-based model (ABM), and a reinforcement learning agent. The Uniswap V3 model provides a representation of the pool, enabling the deployment and interaction with tokens and pools. The ABM introduces complexity by simulating agent interactions and market dynamics, creating a rich environment for strategy evaluation. The reinforcement learning agent, operating within this environment, adopts a deep deterministic policy gradient approach to learn and adapt strategies, aiming for optimal performance in liquidity provisioning. This research aims to develop an intelligent liquidity provisioning (ILP) mechanism using reinforcement learning (RL) to autonomously manage and optimize liquidity within the Uniswap V3 environment. The mechanism seeks to maximize the utility function, considering fees earned, impermanent loss, and other metrics based on liquidity providers’ preferences while adapting to the complex dynamics of the CFMM market. Intelligent Liquidity Provisioning Framework In the RL framework, the liquidity provisioning problem is formulated as a Markov Decision Process (MDP). The MDP consists of states, actions, and rewards. States represent the current market conditions, including asset prices, trading volumes, and other relevant variables. States: Actions correspond to the decisions made by the liquidity provider, such as adjusting liquidity allocations, rebalancing portfolios, etc. Actions: Rewards quantify the desirability of the outcomes based on the liquidity provider’s objective function, preferences, and constraints. The rewards can be positive for desirable outcomes (e.g., high returns) and negative for undesirable outcomes (e.g., high risk or underperformance). Rewards: The objective function represents the liquidity provider’s desired outcome, which can be a combination of factors like maximizing returns, minimizing risks, or achieving a specific trade-off between the two. Constraints can include limitations on liquidity allocations, capital utilization, risk tolerance levels, or other restrictions defined by the liquidity provider. Objective Function: RL training is an iterative process where the agent continuously updates its policy based on feedback. The agent learns from its experiences and refines its decision-making over time, gradually converging to more optimal liquidity provisioning strategies. Once the RL agent has been trained, it can be tested and evaluated using historical data or simulated environments to assess its performance against the liquidity provider’s objective function and constraints. The agent’s performance can be measured using metrics like returns, risk measures, or other relevant performance indicators. By applying RL algorithm, the liquidity provisioning mechanism can learn and adapt to changing market conditions, identify optimal liquidity provision strategies, and balance constraints and preferences specified by the liquidity provider. RL enables the mechanism to find solutions that maximize the liquidity provider’s objective function, considering various trade-offs and constraints autonomously and dynamically. Components of Intelligent Liquidity Provisioning Framework The framework comprises three major components: UniswapV3 Model The Uniswap V3 model implemented in Python offers a detailed and functional simulation of the Uniswap V3 protocol, capturing its nuanced mechanics and providing users with a comprehensive toolset for interacting with the protocol. The UniswapV3_Model class handles the deployment of tokens and pools, initializes pools, and provides an interface for pool actions and pool state retrieval. Overview The Uniswap Model serves as the foundation of the Intelligent Liquidity Provisioning Framework, encapsulating the core mechanics of Uniswap V3. It leverages compiled smart contracts from Uniswap’s V3-Core, deployed in a local Ganache environment using brownie, to create a realistic and interactive simulation. Contract Compilation and Deployment The framework integrates with Brownie, a Python-based development and testing framework for smart contracts, to compile and deploy the Uniswap V3 smart contracts. These contracts are then deployed to a local Ganache environment, providing a sandbox for testing and development. This setup ensures that users can interact with the Uniswap environment without the need for real assets or network transactions, fostering a safe and controlled experimentation space. Agent-Based Simulator is used to simulate the Uniswap V3 environment, agent policies are defined to incorporate the dynamics of Uniswap market participants. Different types of agents are used to simulate the dynamic Uniswap environment Tokenspice agent-based simulator Introduction simulates the actions and interactions of individual agents within the Uniswap V3 ecosystem. By modeling the complex behaviors of different participants, the ABM provides a comprehensive interface of Uniswap V3 dynamic environment, enabling the analysis and optimization of liquidity provisioning strategies. Tokenspice Agent-Based Model (ABM) Agent Types and Behaviors The ABM includes various agent types, each representing a specific role within the Uniswap V3 ecosystem. The two main agents are the Liquidity Provider Agent and the Swapper Agent, which interact with the Uniswap pools to provide liquidity and perform token swaps, respectively. The behavior of these agents is dictated by policies defined in the file, ensuring that their actions are aligned with real-world strategies and market conditions. agents_policies.py This agent adds and removes liquidity from the Uniswap pools. It follows a set of policies that dictate its actions based on the current state of the market and the agent’s preferences. Liquidity Provider Agent: The Swapper Agent performs token swaps within the Uniswap pools, taking advantage of price discrepancies and arbitrage opportunities. Its behavior is guided by policies that assess the potential profitability of trades, considering transaction fees and slippage. Swapper Agent: Simulation Configuration and Execution The file is central to the ABM, configuring how agents interact with each other and with the Uniswap pools. It defines the relationships between agents, policies, and the simulation environment. netlist.py The , , and modules provide the foundational elements for running simulations. The SimEngine orchestrates the simulation, managing the flow of time and the execution of agent actions. The SimStateBase maintains the current state of the simulation, storing data on agent holdings, pool states, and other relevant variables. The SimStrategyBase defines the overarching strategies that guide agent behavior throughout the simulation. SimEngine.py SimStateBase.py SimStrategyBase.py Reinforcement Learning Model Introduction The Reinforcement Learning (RL) Agent is a pivotal component of the Intelligent Liquidity Provisioning Framework, designed to interact with the Uniswap V3 ecosystem through Uniswap Model an agent-based model. This section delves into the RL Agent, its environment, and the DDPG (Deep Deterministic Policy Gradient) algorithm used for training. RL Agent Environment The RL Agent operates in a custom environment, , which interfaces with the Uniswap model and the agent-based model to simulate the DeFi market. This environment facilitates the agent’s interaction with Uniswap pools, allowing it to add and remove liquidity, and observe the consequences of its actions. The RL Agent interacts with the Uniswap model and ABM to simulate real-world liquidity provisioning in Uniswap V3. It chooses actions that result in adding or removing liquidity, with policies and simulation configuration defined in the ABM, ensuring realistic interactions. DiscreteSimpleEnv The environment’s state space includes various market indicators such as the current price, liquidity, and fee growth. These parameters are normalized and provided to the agent at each timestep. State Space: The agent’s action space consists of continuous values representing the price bounds for adding liquidity to a Uniswap pool. These actions are translated into interactions with the Uniswap pools, affecting the state of the environment. Action Space: The reward function is crucial for training the RL Agent. It takes into account the fee income, impermanent loss, portfolio value, and potential penalties, providing a scalar reward signal to guide the agent’s learning process. Reward Function: DDPG Agent The DDPG Agent is a model-free, off-policy actor-critic algorithm using deep function approximators. It can handle high-dimensional state spaces and continuous action spaces, making it well-suited for our Uniswap V3 environment. This network is responsible for providing the best-believed action given a state. It has a sigmoid output layer, outputting relative values for and which are then scaled to the desired range in agent env, representing the price bounds for adding liquidity. Actor Network: price_lower price_upper This network evaluates the action value function, estimating the expected return of taking an action in a given state. Critic Network: DDPG employs target networks for both the actor and critic, which are slowly updated to stabilize training. Target Networks: This technique is used to store a replay buffer of past experiences, allowing the agent to learn from a diverse set of samples, breaking correlations in observations, and smoothing out learning. Experience Replay: Interaction with Uniswap Model and ABM The RL Agent leverages the Uniswap model and agent-based model to simulate real-world liquidity provisioning in Uniswap V3. It interacts with the Uniswap pools through the , performing actions that result in adding or removing liquidity. The agent’s policies and the simulation configuration are defined in the ABM component, ensuring a realistic and coherent dynamic environment. DiscreteSimpleEnv The agent is trained over a series of episodes, each representing a different market scenario (different pool). The agent’s performance is evaluated based on its ability to maximize returns while minimizing risks associated with liquidity provisioning. The effectiveness of the Intelligent Liquidity Provisioning Framework is assessed through the evaluation of the reinforcement learning (RL) agent’s performance. Train and Evaluate Agent: To evaluate the RL agent, we set up a specialized evaluation environment, , which extends the base environment, . This environment is tailored for the evaluation of agent policies. Environment Setup: DiscreteSimpleEnvEval DiscreteSimpleEnv In our evaluation setup, we compare the RL agent’s performance against that of a baseline agent. The baseline agent’s actions are determined by a baseline policy that relies on the current state of the liquidity pool. This agent aims to provide a reference point for evaluating the RL agent’s performance. Baseline Agent: Results Training Evaluation Limitations Currently, the framework does not fully capture the real-time synchronization of pools, which can lead to discrepancies in modeling real Uniswap V3 dynamics. Future work should focus on incorporating mechanisms for better pool synchronization, potentially utilizing tick/position data or events to enhance realism. Pools Synchronization: The agent policies employed in the current framework are relatively simple and naive. To achieve more accurate simulations, future iterations should aim to define more comprehensive agent policies. These policies could model various types of Uniswap agents, such as noise traders, informed traders, retail liquidity providers, and institutional liquidity providers. Alternatively, statistical models trained on historical pool data can inform agent policies for more realistic behavior. Naive Agent Policies: The observation space provided to the agents lacks comprehensive information about the state of the pool. To improve decision-making capabilities, future enhancements should include tick and position data, along with engineered features that offer agents a more comprehensive understanding of the pool’s status. Sparse Observation Space: The action space for agents is currently constrained, with fixed liquidity amounts and restricted price range bounds. Expanding the action space to allow for more flexibility in liquidity provision, as well as considering multiple positions per step, can enhance the fidelity of the simulations. Limited Action Space: Future Work Agent Environment: Implement mechanisms to synchronize pools, possibly using tick/position data or events, to create more realistic dynamics in the Uniswap V3 environment. Synced Pools: Actor/Critic Network Architecture, alpha, beta, tau, batch size, steps, episodes, scaling parameters (rewards, actions, observation space) Hyperparameter Tuning: Define more sophisticated analytical policies that accurately model various Uniswap agents or utilize statistical models trained on historical pool data to inform agent behavior. Comprehensive Agent Policies: Enhance the observation space by including tick and position data, and engineer features that provide agents with a comprehensive view of the pool’s state. Informative Observation Space: Develop an improved reward function that accounts for a wider range of factors, leading to more effective agent training. Improved Reward Function: Instead of one position with a fixed budget at each timestep, implement a more comprehensive mechanism in which the agent is allocated a budget once at the start of the simulation and then learns to use this budget optimally in subsequent steps. Multiple Positions: Define more comprehensive baseline policies to evaluate the performance of the RL agent Baseline Policies: Agent Algorithm Further refine and optimize the hyperparameters of the reinforcement learning agent for better training performance. Hyperparameter Tuning: Explore alternative RL agent models, such as Proximal Policy Optimization (PPO) or Soft Actor-Critic (SAC), to determine if they offer advantages in specific scenarios. Experimentation with Other RL Agents: Investigate the application of multi-agent reinforcement learning techniques, which can be beneficial for modeling interactions among multiple liquidity providers and swappers. Multi-Agent RL (MARL): Implement online learning strategies that allow agents to adapt to changing market conditions in real time, providing a more dynamic and adaptive liquidity provisioning solution. Online Learning: Conclusion In the rapidly evolving landscape of decentralized finance (DeFi), liquidity provisioning plays a pivotal role in enabling efficient and secure trading. Uniswap V3, with its innovative concentrated liquidity feature, has pushed the boundaries of what is possible in DeFi liquidity management. However, the complexities of optimizing liquidity provisioning strategies within this dynamic ecosystem require innovative solutions. Our Intelligent Liquidity Provisioning Framework represents a significant step forward in addressing these challenges. By combining agent-based modeling and reinforcement learning, we have created a powerful toolkit for liquidity providers and market participants. This framework offers a robust and adaptive solution for optimizing liquidity provisioning strategies, with a focus on maximizing utility functions that encompass fees earned, impermanent loss mitigation, and other metrics tailored to individual preferences. Resources Github: https://github.com/idrees535/Intelligent-Liquidity-Provisioning-Framework-V1 References https://kth.diva-portal.org/smash/get/diva2:1695877/FULLTEXT01.pdf https://arxiv.org/pdf/2305.15821.pdf https://github.com/KodAgge/Reinforcement-Learning-for-Market-Making/tree/main https://arxiv.org/ftp/arxiv/papers/2211/2211.01346.pdf https://arxiv.org/pdf/2004.06985.pdf https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9682687 https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0277042 https://deliverypdf.ssrn.com/delivery.php?ID=104119098102026014120072084014107007042068069003049020126088025087121115103007084028042013055035009000054122074096068089064070102052026003014069082076098016080066026088066039027093020006122067093104092065070020126069068106118079127088008098077106031120&EXT=pdf&INDEX=TRUE https://medium.com/blockapex/market-making-mechanics-and-strategies-4daf2122121c https://www.gauntlet.xyz/resources/uniswap-user-cohort-analysis https://gov.uniswap.org/t/uniswap-incentive-design-analysis/21662 https://arxiv.org/pdf/2108.07806.pdf https://www.researchgate.net/publication/341848292_Market_makers_activity_behavioural_and_agent_based_approach https://fruct.org/publications/volume-29/fruct29/files/Struc.pdf https://www.arxiv-vanity.com/papers/1911.03380/ https://insights.glassnode.com/the-week-onchain-week-31-2023/ Also published . here

A Detailed Framework for Intelligent Liquidity Provisioning in Uniswap V3 

Uniswap V3-এ বুদ্ধিমান তারল্য বিধানের জন্য একটি বিস্তারিত কাঠামো

Ein detailliertes Framework für die intelligente Liquiditätsbereitstellung in Uniswap V3

Un marco detallado para el aprovisionamiento inteligente de liquidez en Uniswap V3

Un cadre détaillé pour l'approvisionnement intelligent en liquidités dans Uniswap V3

Uniswap V3 में बुद्धिमान तरलता प्रावधान के लिए एक विस्तृत रूपरेखा

Uniswap V3 におけるインテリジェントな流動性プロビジョニングの詳細なフレームワーク

Uniswap V3의 지능형 유동성 프로비저닝을 위한 세부 프레임워크

Uma estrutura detalhada para provisionamento inteligente de liquidez no Uniswap V3

Подробная структура интеллектуального предоставления ликвидности в Uniswap V3

Uniswap V3'te Akıllı Likidite Tedariği için Ayrıntılı Bir Çerçeve

Khung chi tiết để cung cấp thanh khoản thông minh trong Uniswap V3

Uniswap V3 智能流动性配置的详细框架

Computer vision is a field experiencing significant growth and has numerous practical applications, ranging from self-driving cars to facial recognition systems. However, one of the major challenges in this field is obtaining high-quality datasets to train machine learning models. To address this challenge, torchvision provides access to pre-built datasets, models and transforms specifically designed for computer vision tasks. Torchvision also supports both CPU and GPU acceleration, making it a flexible and powerful tool for developing computer vision applications. What are “Torchvision Datasets”? Torchvision datasets are collections of popular datasets commonly used in computer vision for developing and testing machine learning models. With torchvision datasets, developers can train and test their machine learning models on a range of tasks, such as image classification, object detection, and segmentation. The datasets are also preprocessed, labeled and organized into formats that can be easily loaded and used. List of the Torchvision Datasets MNIST CIFAR-10 CIFAR-100 ImageNet COCO Fashion-MNIST SVHN STL-10 CelebA Pascal VOC Places365 1. MNIST This torchvision dataset is popular and widely used in the fields of machine learning and computer vision. It consists of 70,000 grayscale images of handwritten digits 0–9, with 60,000 images for training and 10,000 for testing. Each image is 28x28 pixels in size and has a corresponding label denoting which digits it represents. To access this dataset, you can download it directly from or load the dataset using torchvision: Kaggle import torchvision.datasets as datasets
# Load the training dataset
train_dataset = datasets.MNIST(root='data/', train=True, transform=None, download=True)
# Load the testing dataset
test_dataset = datasets.MNIST(root='data/', train=False, transform=None, download=True) Code for loading MNIST dataset using PyTorch torchvision package. Retrieved from . on 20/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.MNIST.html#torchvision.datasets.MNIST 2. CIFAR-10 The CIFAR-10 dataset consists of 60,000 32x32 colour images in 10 classes, with 6,000 images per class. It has a total of 50,000 training images and 10,000 test images which is further divided into five training batches and one test batch, each with 10,000 images. This dataset can be downloaded from , or loaded here using torchvision: Kaggle import torch
import torchvision
import torchvision.transforms as transforms
 
transform = transforms.Compose(
    [transforms.ToTensor(),
     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
 
trainset = torchvision.datasets.CIFAR10(root='./data', train=True,
                 	                   download=True, transform=transform)
testset = torchvision.datasets.CIFAR10(root='./data', train=False,
                                       download=True, transform=transform)
 
trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,
                                          shuffle=True, num_workers=2)
testloader = torch.utils.data.DataLoader(testset, batch_size=4,
                                         shuffle=False, num_workers=2) Note that you can adjust the batch size and number of worker processes for the data loaders as needed. Code for loading CIFAR-10 dataset using PyTorch torchvision package. Retrieved from . on 20/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.CIFAR10.html#torchvision.datasets.CIFAR10 3. CIFAR-100 The CIFAR-100 dataset has 60,000(50,000 training images and 10,000 test images) 32x32 colour images in 100 classes, with 600 images per class. The 100 classes are grouped into 20 super-classes, with a fine label to denote its class and a coarse label to represent the super-class that it belongs to. To download the torchvision dataset from Kaggle, please visit the Kaggle and follow the instructions provided there. Alternatively, if you prefer to load the dataset using the torchvision library, you can do this: website import torchvision.datasets as datasets
import torchvision.transforms as transforms
# Define transform to normalize data
transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
])
 
# Load CIFAR-100 train and test datasets
trainset = datasets.CIFAR100(root='./data', train=True, download=True, transform=transform)
testset = datasets.CIFAR100(root='./data', train=False, download=True, transform=transform)
 
# Create data loaders for train and test datasets
trainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)
testloader = torch.utils.data.DataLoader(testset, batch_size=64, shuffle=False) Code for loading CIFAR-100 dataset using PyTorch torchvision package. Retrieved from . on 20/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.CIFAR100.html#torchvision.datasets.CIFAR100 4. ImageNet The ImageNet dataset in torchvision contains approximately 1.2 million training images, 50,000 validation images and 100,000 test images. Each image in the dataset is labeled with one of the 1,000 categories such as "cat," "dog," "car,", "airplane" etc. To download this torchvision dataset, you have to visit the or load in torchvision: website import torchvision.datasets as datasets
import torchvision.transforms as transforms
 
# Set the path to the ImageNet dataset on your machine
data_path = "/path/to/imagenet"
 
# Create the ImageNet dataset object with custom options
imagenet_train = datasets.ImageNet(
	root=data_path,
	split='train',
    transform=transforms.Compose([
        transforms.Resize(256),
        transforms.RandomCrop(224),
        transforms.RandomHorizontalFlip(),
        transforms.ToTensor(),
        transforms.Normalize(
            mean=[0.485, 0.456, 0.406],
            std=[0.229, 0.224, 0.225])
	]),
	download=False
)
 
imagenet_val = datasets.ImageNet(
	root=data_path,
	split='val',
    transform=transforms.Compose([
        transforms.Resize(256),
        transforms.CenterCrop(224),
        transforms.ToTensor(),
        transforms.Normalize(
            mean=[0.485, 0.456, 0.406],
            std=[0.229, 0.224, 0.225])
	]),
	download=False
)
 
# Print the number of images in the training and validation sets
print("Number of images in the training set:", len(imagenet_train))
print("Number of images in the validation set:", len(imagenet_val)) Code for loading ImageNet dataset using PyTorch torchvision package. Retrieved from . on 21/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.ImageNet.html#torchvision.datasets.ImageNet 5. MS Coco The Microsoft Common Objects in Context(MS Coco) dataset contains 328,000 high-quality visual images of everyday objects and humans, often used as a standard to compare the performance of algorithms in real-time object detection. To download this torchvision dataset, please visit the or load in torchvision: website import torch
from torchvision import datasets, transforms
# Define transformation
transform = transforms.Compose([
   transforms.Resize(256),
   transforms.CenterCrop(224),
   transforms.ToTensor(),
   transforms.Normalize(mean=[0.485, 0.456, 0.406],
                        std=[0.229, 0.224, 0.225])
])
 
# Load training dataset
train_dataset = datasets.CocoDetection(root='/path/to/dataset/train2017',
                                       annFile='/path/to/dataset/annotations/instances_train2017.json',
                                       transform=transform)
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=32, shuffle=True)
 
# Load validation dataset
val_dataset = datasets.CocoDetection(root='/path/to/dataset/val2017',
                                     annFile='/path/to/dataset/annotations/instances_val2017.json',
                                     transform=transform)
val_loader = torch.utils.data.DataLoader(val_dataset, batch_size=32, shuffle=False) Make sure to replace the /path/to/dataset placeholders with the actual path to your dataset directory. Also, adjust the batch_size parameter to fit your needs. Code for loading MS Coco dataset using PyTorch torchvision package. Retrieved from . https://pytorch.org/vision/stable/generated/torchvision.datasets.CocoDetection.html#torchvision.datasets.CocoDetection on 21/3/2023 6. Fashion-MNIST The Fashion MNIST dataset was created by Zalando Research as a replacement for the original MNIST dataset. The Fashion MNIST dataset consists of 70,000 grayscale images (training set of 60,000 and a test set of 10,000) of clothing items. The images are 28x28 pixels in size and represent 10 different classes of clothing items, including ­T-shirts/tops, trousers, pullovers, dresses, coats, sandals, shirts, sneakers, bags, and ankle boots. It is similar to the original MNIST dataset, but with more challenging classification tasks due to the greater complexity and variety of the clothing items. This torchvision dataset can be downloaded from , or loaded using this code: Kaggle import torch
import torchvision
import torchvision.transforms as transforms

# Define transformations
transform = transforms.Compose(
    [transforms.ToTensor(),
     transforms.Normalize((0.5,), (0.5,))])
# Load the dataset
trainset = torchvision.datasets.FashionMNIST(root='./data', train=True,
                                        download=True, transform=transform)

testset = torchvision.datasets.FashionMNIST(root='./data', train=False,
                                       download=True, transform=transform)

# Create data loaders
trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,
                                          shuffle=True, num_workers=2)

testloader = torch.utils.data.DataLoader(testset, batch_size=4,
                                         shuffle=False, num_workers=2) Code for loading Fashion-MNIST dataset using PyTorch torchvision package. Retrieved from . on 21/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.FashionMNIST.html#torchvision.datasets.FashionMNIST 7. SVHN The SVHN (Street View House Numbers) dataset is an image dataset derived from Google’s Street View imagery, which consists of cropped images of house numbers taken from street-level images. It is available in a full format with all the house numbers and their bounding boxes and a cropped format with just the house numbers alone. The full format is often used for object detection tasks, while the cropped format is commonly used for classification tasks. The SVHN dataset is also included in the torchvision package and it contains 73,257 images for training, 26,032 images for testing and 531,131 additional images for extra training data. To download this torchvision dataset, you can go to or you can load it here: Kaggle import torchvision
import torch

# Load the train and test sets
train_set = torchvision.datasets.SVHN(root='./data', split='train', download=True, transform=torchvision.transforms.ToTensor())
test_set = torchvision.datasets.SVHN(root='./data', split='test', download=True, transform=torchvision.transforms.ToTensor())

# Create data loaders
train_loader = torch.utils.data.DataLoader(train_set, batch_size=64, shuffle=True)
test_loader = torch.utils.data.DataLoader(test_set, batch_size=64, shuffle=False) Code for loading SVHN dataset using PyTorch torchvision package. Retrieved from . on 22/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.SVHN.html#torchvision.datasets.SVHN 8. STL-10 The STL-10 dataset is an image recognition dataset that consists of 10 classes, with a total of about 6,000+ images. The STL-10 stands for “Standard Training and Test Set for Image Recognition-10 classes” and the 10 classes in the dataset are: Airplane Bird Car Cat Deer Dog Horse Monkey Ship Truck To access this dataset, you can download it directly from or load the dataset using torchvision: Kaggle import torchvision.datasets as datasets
import torchvision.transforms as transforms

# Define the transformation to apply to the data
transform = transforms.Compose([
    transforms.ToTensor(),   

# Convert PIL image to PyTorch tensor
    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))   # Normalize the data
])

# Load the STL-10 dataset
train_dataset = datasets.STL10(root='./data', split='train', download=True, transform=transform)
test_dataset = datasets.STL10(root='./data', split='test', download=True, transform=transform) Code for loading STL-10 dataset using PyTorch torchvision package. Retrieved from . on 22/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.STL10.html#torchvision.datasets.STL10 9. CelebA This torchvision dataset is a popular large-scale face attributes dataset which comprises over 200,000 celebrity images. It was first released by researchers at the Chinese University of Hong Kong in 2015. An image in the CelebA consists of 40 facial attributes such as age, hair colour, facial expression and gender. Also, these images were retrieved from the internet and cover a wide range of facial appearances, including different races, ages and genders. Bounding box annotations for the location of the face in each image, as well as 5 landmark points for the eyes, nose and mouth. You can download this dataset on or load it using this code: Kaggle import torchvision.datasets as datasets
import torchvision.transforms as transforms

transform = transforms.Compose([
    transforms.CenterCrop(178),
    transforms.Resize(128),
    transforms.ToTensor(),
    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))
])
celeba_dataset = datasets.CelebA(root='./data', split='train', transform=transform, download=True) Code for loading CelebA dataset using PyTorch torchvision package. Retrieved from . on 22/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.CelebA.html#torchvision.datasets.CelebA 10. PASCAL VOC The VOC dataset (Visual Object Classes) was first introduced in 2005 as part of the PASCAL VOC Challenge, which aimed at advancing the state of the art in visual recognition. It consists of images of 20 different object categories, including animals, vehicles, and common household objects. Each of these images is annotated with the locations and classifications of objects within the image. The annotations include both bounding boxes and pixel-level segmentation masks. The dataset is split into two main sets: the training and validation sets. The training set contains approximately 5,000 images with annotations, while the validation set contains around 5,000 images without annotations. In addition, the dataset also includes a test set with approximately 10,000 images, but the annotations for this set are not publicly available. To access the recent dataset, you can download from the , or load it in torchvision: website Kaggle import torch
import torchvision
from torchvision import transforms

# Define transformations to apply to the images
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

# Load the train and validation datasets
train_dataset = torchvision.datasets.VOCDetection(root='./data', year='2007', image_set='train', transform=transform)
val_dataset = torchvision.datasets.VOCDetection(root='./data', year='2007', image_set='val', transform=transform)
# Create data loaders
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=32, shuffle=True)
val_loader = torch.utils.data.DataLoader(val_dataset, batch_size=32, shuffle=False) Code for loading PASCAL VOC dataset using PyTorch torchvision package. Retrieved from . on 22/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.VOCDetection.html#torchvision.datasets.VOCDetection 11. Places365 The Places365 dataset is a large-scale scene recognition dataset with more than 1.8 million images covering 365 scene categories. The Places365 Standard dataset consists of around 1.8 million images, while the Places365-Challenge dataset contains 50,000 additional validation images that are more challenging for recognition models. To access this dataset, you can use or load in torchvision here: Kaggle import torch
import torchvision
from torchvision import transforms

# Define transformations to apply to the images
transform = transforms.Compose([
    transforms.Resize((224, 224)),
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
])

# Load the train and validation datasets
train_dataset = torchvision.datasets.Places365(root='./data', split='train-standard', transform=transform)
val_dataset = torchvision.datasets.Places365(root='./data', split='val', transform=transform)
# Create data loaders
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=32, shuffle=True)
val_loader = torch.utils.data.DataLoader(val_dataset, batch_size=32, shuffle=False) Code for loading Places365 dataset using PyTorch torchvision package. Retrieved from . on 22/3/2023 https://pytorch.org/vision/stable/generated/torchvision.datasets.Places365.html#torchvision.datasets.Places365 Common Use Cases for Torchvision Datasets - This dataset is commonly used for image classification tasks, particularly for handwritten digit recognition. MNIST - These datasets are commonly used for image classification tasks, particularly for object recognition. and CIFAR-10 CIFAR-100 - This torchvision dataset contains millions of images of various objects and animals and is commonly used for object recognition tasks. ImageNet - This dataset is commonly used for object detection, segmentation, and captioning tasks. MS COCO - Similar to MNIST, this dataset is commonly used for image classification tasks, particularly for fashion item recognition. Fashion-MNIST - This dataset is commonly used for digit recognition tasks. It is particularly useful for developing and testing algorithms for recognizing numbers in real-world settings. SVHN - This torchvision dataset is commonly used for image classification tasks, particularly for object recognition. STL-10 - This dataset is commonly used for facial recognition and attribute classification tasks. It consists of images of various facial expressions, poses and backgrounds. CelebA - This dataset is commonly used for object detection and segmentation tasks. Pascal VOC - This dataset is commonly used for scene recognition tasks. It consists of images of various indoor and outdoor scenes such as bedrooms, beaches, and offices. Places365 Final Thoughts Torchvision datasets are often used for training and evaluating machine learning models such as convolutional neural networks (CNNs), which are commonly used in computer vision applications. They are also available for anyone to download and use freely. The lead image of this article was generated via model using the prompt 'thousands of images organized together in small frames'. HackerNoon's AI Stable Diffusion More Dataset Listicles: Excel Datasets Keras Datasets R Datasets PyTorch Datasets Hugging Datasets

11 Torchvision Datasets for Computer Vision You Need to Know

কম্পিউটার ভিশনের জন্য 11 টর্চভিশন ডেটাসেট আপনার জানা দরকার

11 Torchvision-Datensätze für Computer Vision, die Sie kennen müssen

11 conjuntos de datos de Torchvision para visión por computadora que necesita saber

11 ensembles de données Torchvision pour la vision par ordinateur que vous devez connaître

कंप्यूटर विजन के लिए 11 टॉर्चविजन डेटासेट आपको जानना जरूरी है

知っておくべきコンピューター ビジョン用の 11 の Torchvision データセット

당신이 알아야 할 11가지 컴퓨터 비전용 Torchvision 데이터 세트

11 conjuntos de dados Torchvision para visão computacional que você precisa conhecer

Bilgisayarlı Görme için Bilmeniz Gereken 11 Torchvision Veri Kümesi

11 bộ dữ liệu Torchvision cho thị giác máy tính bạn cần biết

你需要知道的 11 个用于计算机视觉的 Torchvision 数据集

ARXVISION

arx.vision

1-10 employees

Since 2016

COMPANY RANKING

ARXVISION

EVERGREEN INDEX #1214

Representing this company? Click here to claim and customize this page!

ARxVision WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

ARXVISION

arx.vision

1-10 employees

Since 2016

COMPANY RANKING

ARXVISION

EVERGREEN INDEX #1214

RELATED COMPANIESRANK

STORIES ABOUT ARXVISION

Representing this company? Click here to claim and customize this page!

ARxVision WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!