Let's Enhance is a deep tech AI company that builds End-to-End imaging solutions for photo enhancement and optimization

Let's Enhance

letsenhance

5 Effective Methods for Passing Parameters to Apache APISIX

The Ultimate React State Management Handbook

Everyone knows that project-based learning is best. Endless ‘diving-in’ tutorials might be good initially but if you can’t build a project all by yourself, you have not learned much. Real learning requires practice and whether you want to improve your skills or build a portfolio to get a programming job, the best way to practice is to build projects. Now, what projects you should build is another good question that I will try to answer in this article. While there’s no universal answer to what kind of projects one should build when learning to code, there are definitely some good project categories and ideas that we are going to discuss in this article. These projects will help you upskill yourself and create a portfolio of projects as a web developer or mobile app developer that you can show to your future employers in interviews. So without further ado, let’s get started: Content Overview Web portfolio projects Landing page API-powered website Clone website CRUD application AI-powered website Full stack  unique project Mobile app portfolio projects Ecommerce app Diet tracking app Personal finance app Movie review app Photo editing app Social networking app Wrapping up Web portfolio projects 1. Landing page All great products on the internet have their own landing page. As a web developer, knowing how to build great landing pages is the most basic thing. This would require you to know how to build responsive websites and websites that have good design. Landing pages are static so even if you just know and , you can come up with beautiful landing pages that demonstrate the product best. You can browse on Frontend Mentor for landing page ideas. HTML CSS 2. API powered website Most projects consume some API and if you are serious about building a career as a programmer, you must know how to consume APIs to build better projects. There are endless ideas that you can make real with APIs. From fun Pokémon projects to Music streaming apps to AI projects(which will be discussed later in the article), your creativity is the only limit to what you can build. 3. Clone website A Clone website could be a copy of any popular websites out there or the ones that you like and use the most. It could be anything like Facebook, , , , , etc., or even the not-so-popular ones given that they are somewhat complex to build. YouTube Instagram Twitter Amazon If you can create a clone of such websites, it demonstrates that you can convert good designs into good code. Find what website you want to clone and start building. If you want to take it a step further, create a full-stack clone application with a frontend, API, and database. 4. CRUD application Most websites have CRUD functionality enabled. CRUD operations allow users to create, read, update, and delete data. Blogs, e-commerce, social media, educational platforms, productivity apps, etc. all have some sort of CRUD functionality enabled. There are lots of good project ideas under this category like a TO-DO list, Notes App, Chat App, E-Commerce, Social Media website, etc. that you can build. 5. AI-powered website AI is booming in 2023 and the best part is that you don’t necessarily need advanced artificial intelligence knowledge to build AI projects. You can build something great on top of great AI APIs available on the internet like Open AI API etc. Having an AI project in your portfolio is a great step to show your employers that you care about the latest technological developments and can make your way to learning the required skills. 6. Full Stack unique project This is the sum total of all points we have discussed above. You have built landing pages, API-powered web apps, Clone web apps, CRUD apps, and maybe AI-powered web apps. Now, you have to combine all this and build a unique project. Here you will create a fully functional full-stack project along with a separate landing page. If you can also add AI to it, that would be an extra bonus. This project should ideally be something you can relate to, an idea you find interesting, or a solution to a problem that you are facing. Once you have finalized the idea, brainstorm and document all the features that you want to add to the web app. Now, you should start the building process, there are three steps to it: Create all the API endpoints you need and check if the data is getting saved properly in the database or not. Frontend: This is where all the users of your web app will land. So try to build the app with good UI/UX that users would love to use. API endpoints: The landing page doesn’t add anything to your product but it’s for the people who will come to use your project. So make sure, you spend a good time on it. If you can build a big project like this nicely, it would serve as the best certification of your skills. Landing Page: Mobile portfolio projects We have already discussed web projects and most of those project ideas can also be used in the context of apps like building a clone app, CRUD app, AI app, etc. But for this section, we are going to discuss specific project examples so that after you finish reading this article, you will have both specific ideas and broad categories that you can fit your projects under. 1. ECommerce app ECommerce is rocking both on the web and mobile and we can’t ignore the massive size of this industry. Knowing how to build and maintain e-commerce apps is going to help you a lot as a mobile app developer. To begin with, you can either clone an existing e-commerce app like Amazon, etc or you can build your own from scratch. Irrespective of what you choose, let’s see some key features that you must add to your app: Search for a product Add to cart and checkout Add your own products Past purchases Rate and review products 2. Diet tracker app Diet Tracker app is an app with lots of potential. You can start this as a simple food logging app and eventually slowly make it bigger and better. Let’s see some features that you can add to this app, sooner or later: Food logging Find nutrients by searching the name of a recipe Diet plan for specific fitness goals Weight tracking Water reminder Progress tracking 3. Personal finance app Managing finances is a gruesome task and what’s a better way to simplify it than with an app? Let’s discuss some features this app could have: Record income and expenses Different categories for income and expenses Can give tags to a transaction Budget planning Charts and graphs to view the cash flow, income, and expenses Can automate monthly income and fixed expenses Track your investments While there’s no end to what you can add to this app, even if you implement the above features, that would make this app a solid project in your portfolio. 4. Movie review app Apps like IMDb are really popular and fun to build. You have to create a similar app in this section and make sure that you definitely add the following features to it: Search for a movie Rate a movie Write review Read existing reviews Follow tags for various genres Read the description, cast, release date, budget, collection, etc. While these features would be interesting to implement, don’t forget to differentiate yourself wherever you can. For instance, you can create a niche movie app like Sci-Fi, Comedy, Action, etc. 5. Photo editing app Unlike other apps mentioned in this section, the Photo editing app is a little different and it will help to diversify the project categories in your portfolio. So, let’s see some features that you can add to the project: Enhancement options like contrast, brightness, etc. Create a collage Add borders and frames Filters like Grayscale, Lighten, Darken, etc. Reduce file size Remove background AI enhancement 6. Social networking app Social Networking apps are everyone’s favorite and have a bright future. Building one such project will give your skills and portfolio great improvement and help you shine in front of your employer. Let’s see some features that you can add to your app: Create posts with text and images Comment on other people’s posts Like/Upvote and Bookmark posts Follow other people based on interests Personalized feed AI-generated posts and images The AI-generated posts and images are just an additional feature to spark up things that you can consider adding while the rest of the features are basic and you must add them to your project. Wrapping up That’s it. I hope you are filled with endless ideas for your upcoming projects. You have both idea categories and specific examples that you can build whether you are a web developer, mobile app developer or both. Make sure to spend good time on the user interface and user experience of these projects as they should be good enough to be used by real user. Moreso, a great design and user experience is the key to acquiring more users to your projects/apps. I hope you enjoyed this article, let me know which project are you going to choose to start your journey. Good luck! Also published here.

12 Project Ideas for Aspiring Web and Mobile Developers

The Benefits of AI for Artists While many are still divided on the topic of AI art, we are firm believers that generative AI can offer unique benefits to artists. New technology has improved the creative process throughout history, and here’s how generative AI does the same. Takeaways AI-based art tools make new creative skills accessible to those without the time or money to invest in learning them from the ground up; Generative AI can offer design concepts to help artists overcome their creative blocks; AI is capable of enhancing art in ways that are close to impossible to do by hand; Conceptualizing new creative projects is made easy with the help of AI, reducing the time it takes to create and finish a new piece. Accessibility Blender is one of the best-known 3D modeling and animation software to work with. It’s by no means the first, but in recent years, it’s gotten really popular for one very good reason: it’s accessible in more ways than one. First of all, it’s completely free. So long as you have hardware strong enough to handle it, you can download it and start sculpting or animating. Secondly, it’s accessible thanks to a user-friendly interface that is very easy to learn and use. With the release of Blender, thousands of new artists across the globe gained access to 3D sculpting and animation tools that helped them create new art and start their careers as artists. Now, this doesn’t directly translate into generative AI, since rendering an image with Midjourney doesn’t make someone an illustrator or painter. However, the accessibility in the case of AI art is about giving artists more opportunities to make useful additions to their work. Here’s a great example of how accessible AI tools can help artists overcome challenges in their work. A comic book illustrator may be very skilled in designing characters but is unable to draw gorgeous backgrounds. Normally, they’d have two options: either take time to learn the skill themselves or commission another artist to help them. In the case of artists who are just starting out, dedicating time and money to either of these solutions may not be an option. With the help of generative AI, this artist can easily create a background for their comic. There will always be demand for skilled background artists or artists of any kind for that matter. But for those who don’t have the privilege of learning the skill themselves or the funding to pay for a professional, generative AI is a useful alternative. Creating a Manga Background Let’s put our aforementioned example of a comic book artist to the test with and create a manga background. Midjourney and . Register on Discord create a private server Go to the Midjourney website and join the Beta. After you’ve joined, Midjourney will ask you to join the official Midjourney Discord server. There, you can generate images, however, we still need to add the Midjourney bot to our own private Discord server. In the official Midjourney Discord server, on the right-hand side of the user interface, find the Midjourney bot, click on it, and add it to your private server. In any text channel in your private server, type in followed by whatever it is you’d like the AI to generate. /imagine After a few seconds, you should see the results of your prompt in the text channel of your Discord server. Click on any of the 4 options, and right-click to download. As you can see, Midjourney was able to create vibrant backgrounds that artists can use as inspiration on how to draw their own. Pro Tip: There are tricks to writing effective text prompts for AI image rendering. on how to create prompts for image generators. Check out this in-depth guide Inspiration Even artists with decades of experience in their craft are sometimes at a loss for ideas. Artist’s block is very common, where artists reach a temporary creative limit. No words come to mind, no ideas on where the brushstroke should go, or what final note to add to the composition to complete it. This does not mean the artist is incapable of creative work, but it is the result of an infinite number of circumstances, such as emotional exhaustion, lack of motivation, interruptions from personal matters, etc. Generative AI has the potential to provide artists with the inspiration they need to overcome artist's block and continue developing their creative work. Let’s take writing as an example. If I had writer’s block while writing this article, I could use platforms like ChatGPT or Lex to help me understand what elements are missing from the work that would allow me to continue writing. They could also provide me with new wording or paragraph structuring ideas that would help me avoid writer’s block and continue working. It’s important to keep in mind, that generative AI functions as a database for art. This means that if you’re looking for something new, quick use of generative AI could assist you in discovering new artwork and creative pieces that will provide you with the inspiration you need to continue working on your own stuff. Think of it like visiting a museum with an infinite number of paintings or reading excerpts from hundreds of authors, each with their unique voice. Overcoming Creative Block for Character Design Let’s use Midjourney once more, and this time, to generate concepts for a Steampunk character. Much like with the previous Midjourney example, you’re going to need a Midjourney bot that has been invited to your private server. Type in the text prompt for your desired character concept. For this use case, we wrote a steampunk woman with goggles wearing a mask; Here are the results of our text prompt; Now, you can use these rendered images as inspiration for your own character designs. Because the AI targets specific features that are indicative of the steampunk style, you can replicate these prominent elements in your own artwork and grasp the fundamentals of this art style. Enhancement We’ve been using software to enhance our work for some time now. Software that automatically applies filters to photos adjusts lighting in videos and enhances the quality of images has existed for quite a while. This technology has made it easier for artists to enhance their work, and AI generation can help in this regard as well. While still in the early stages of development, there are AI platforms specialized in turning a static drawn image into an animated GIF. Prior to the development of such technologies, artists had to draw and animate these on their own. Now, an artist who does not possess these skills can feature additional enhancements and make their art more complex. The same goes for lighting, adding filters, unique stylizations, and so on. This is especially useful for artists working on a very urgent project who do not have the time to do all this by hand. But with generative AI, they have more options to enhance their work in a very short amount of time. Upscaling a Photo A great example of how AI can enhance art is by upscaling it. Let’s Enhance is an AI-powered platform that can upscale and increase the resolution of images without loss in quality, something that is pretty much impossible to do without the help of AI. Create an account on . Letsenhance.io Click on and you will arrive at the photo editing dashboard. My Images, Upload the image, and tweak the options on the right-hand side to the image type and by how much you’d like the image to be upscaled. Click and once that’s done, save the image to your device. We upscaled one of the images from a prior example, which was 446 by 442 pixels to 1784 by 1768 pixels. Here’s a side-by-side comparison of the two images. Start processing, Pro Tip: Check out this on how to upscale your images. complete guide Conceptualizing Creating art is a procedural process for many artists, where they create a prototype or first draft of their work, then keep chipping away or adding new portions to it and create a complete piece. This process is a crucial part of creation as makes the final product more complex and dynamic. The challenge of working on WIPs is that artists spend days, even weeks creating a concept for a new design, only to have it be rejected. For design teams working with very strict deadlines, generative AI can help them develop tons of design iterations, present them to clients, and then start developing the WIPs that are approved. This means greater results in a shorter amount of time, which benefits both artists, who now don’t need to worry as much about creating everything from scratch, and also clients, who now receive exactly what they want without having to wait for the commissioned artist to create new concepts. Storyboarding a Film Using Dall.E 2, we tried to see how well AI could help us create a storyboard and conceptualize how a film will look before a single frame is even shot. Register an account on . Write the text prompt in the designated space. Make sure to include stylistic keywords, such as and and are just a few examples of keywords that are very useful for brainstorming film frames. openai.com/product/dall-e-2 black and white, noir, wide shot, photo. Wide shot, close-up, over-the-shoulder 2. Once the images are rendered, hover over the images to download the one that you think fits the style of your film. Using AI, scriptwriters can visualize their film as they are writing the script, reducing the time it takes to conceptualize the storyboard. Concluding Words It’s important to understand that generative AI is still in its infancy, and it’s far too early to tell which direction it will go, both in terms of technology and legislation. It may just be that in the near future, policies and regulations are adopted to limit the use of generative AI to ensure that human creativity is not curtailed. It’s hard to tell, but in an ever-evolving world of technology, it’s important for everyone to learn and adapt to the changing landscape of the creative field. Computer-generated imagery (CGI) has taken off recently to the point where anyone with half-decent hardware can create their own animated films. But skilled 2D animators are still in high demand. AI was able to write content all the way back in the 1980s, and yet, writing is still a bustling industry. Creating art by hand did not disappear with the introduction of digital tools like Photoshop. AI is not here to replace artists. Instead, much like with any technological advancement, it is here to offer convenience and make the creative process more practical for those who are looking for efficiency in creative designs. And while there is still plenty to discuss regarding the ethical use of AI art, it’s important to acknowledge how this new technology can benefit artists and their hard work.

How Generative AI Can Boost Creativity and Enhance Art

174 AI Tools To Try

This guide will cover the basics of working with text prompts and how you can effectively utilize them to render the images you envisioned. Takeaways Prompts for AI image generators can be both text and image-based, and in some cases, a mix of both The general structure of text prompts for AI image generators is similar across the board, as is the word count and order of descriptors. Different elements of a text prompt have unique effects on the rendering and add complexity to the image. What Are Prompts? Prompts for AI image generation are commands, either textual or visual, that dictate what image the AI platform creates. Typically, you will find two different image generation prompt types: text and image. Text Prompts To tell the AI what kind of image to render, you should first describe that image. This is done with a text prompt, which is a list of words, sometimes even a grammatically correct sentence, describing what the AI should render. The highlighted textbox is an example text prompt used to . Depending on the rendering platform you’re using, identical text prompts will have varying results. render images with AI using the Let’s Enhance Image Generator Image Prompt In some cases when text is not enough, you can upload images to an AI image generator and use it as a reference to create other images. Depending on the type of AI platform you’re using and the type of results you will need, using image prompts can be even more effective than text. For example, DALL.E’s Outpainting AI can take an existing image and render its continuation. The image above demonstrates what the AI believes is the continuation of the famous Johannes Vermeer painting of Girl with a Pearl Earring. The original image of the painting is within the square and everything that surrounds it was rendered by DALL.E. A Mix of Both While most of the AI image generators you find on the market today will be text-based with some options for image prompts, there are ways to utilize a mix of both prompts to render desired images. Writing an Image Generation Prompt One very important detail you must keep in mind is to understand what a particular AI image generator specializes in and how to use it effectively. Some AI image generators, such as the Dall-E series are very good at rendering images that imitate photorealism. On the other hand, there’s Midjourney, which is well known for being very effective in digital art or illustration formats. Our own Let’s Enhance Image Generator is great at simulating the visual aesthetic of product images, on top of also rendering photo-realistic images, illustrations, and 3D models. Word Count: Use At Least 3-7 Words While most AI image generators will have a character limit on the text prompts, there is no word limit. This means that you have the freedom to structure your text prompt any way you want. However, the consensus, for now, seems to be that you need 3-7 words to write a successful text prompt for the AI to render. Here are the rendered images of the text prompt in the example above. The “wide shot” prompt significantly impacted the images in the left column. You don’t have to strictly abide by this rule and are allowed to use less than 3 or more than 7 words. But as you’ll see in the examples below, it’s not so much the number of words that matter, but what words you use. Subject: Who and What Any AI image render requires some sort of subject. This should be the first thing you write in the text prompt (in most cases). It is possible to just leave a single subject in the text prompt and have the AI generate something a little more general. However, in most cases, you’re going to need a more descriptive text prompt to generate an image that comes as close to what you were looking to render. Subjects are typically nouns, so the best place to start with your text prompt regardless of which AI you’re utilizing is to start with one. Description: What They Are Doing, Where, and How To add a little more nuance to your rendered image and help the AI narrow down what images to use as references, you need to utilize descriptions. These are your adjectives, and much like how you would use them in any ordinary sentence, they describe the subject. The trick with using adjectives in AI image generation is that you can combine several of these to add more depth and complexity to the image. Aside from adjectives that describe the subject, another element of the description is details that answer questions such as: What is happening? What is the subject doing? How is the subject doing this? What’s happening around the subject? That last question is pretty important, as backgrounds play a key role in effective descriptions. Here’s an example of two similar text prompts yielding quite different results. The text prompt for the image on the left was . Raccoon reading The text prompt for the right image, however, was slightly more detailed; hence why the image came out clearer and more complex: Raccoon reading a book in a library photo close shot. As you can see by the results, we answered several questions, and the AI was able to narrow down what we were looking for and rendered something as close to what we were looking for as possible. The AI was able to generate an image with greater details and fewer anomalies. Here’s another great example of how AI image generators work better with more detailed image prompts. This generated image of a finch demonstrates how AI can render a generic but very accurate image even without additional descriptors. However, the rendering on the right demonstrates how AI image generators, in this case, Stable Diffusion, can work with more elements to render a significantly more complex visual. Aesthetic and Style: How It Looks To complete the text prompt, there are a few more text prompts you must add that will put the finishing touches on your rendered image. Notice how, at the end of the text prompt above, we had words like “photo” and “close up”. The last words in your prompt will be the ones that dictate the overall aesthetic and style of your image. Instead of “photo”, you could have had “oil painting”, “fresco”, or “3D rendering”. To dictate the style of your image, you can also use art movements such as impressionist, gothic, steampunk, etc. Also, if you’re looking to replicate the works of a famous artist, you can even put one or more famous names in the text prompt too. The AI will combine influences from various famous artists to create something that blends elements from all of them. Finally, you can also dictate the framing of the image as well. Prompts such as “close up” or “medium shot”, much like how you would do in photography, can help the AI understand what angle you want the image to be presented from and from what distance. Here’s an example of two images generated by Lexica of a handheld computer device. The prompt on the left, though simple, was able to generate a vintage device that even replicates the old-school LCD display. However, the rendering to the right, thanks to prompts like “vaporwave aesthetic” and “product photography”, has a more defined visual aesthetic and style, as you can tell by the neon colors. Experience Is the Best Teacher No amount of tips and tricks are going to teach you how to write effective AI prompts than experience. The best part about AI image generators is that they’re very accessible. All you need is an internet connection, and you can choose from the many free platforms available. So get out there and experiment with AI image prompts to see what new and fascinating things you can discover in the rising field of image generation.

A Simple Guide on AI Text Prompts: Tell the AI Exactly What You Want to See

এআই টেক্সট প্রম্পটগুলির একটি সহজ নির্দেশিকা: আপনি কী দেখতে চান তা এআইকে বলুন

Eine einfache Anleitung zu KI-Textaufforderungen: Sagen Sie der KI genau, was Sie sehen möchten

Una guía simple sobre mensajes de texto de IA: dígale a la IA exactamente lo que quiere ver

Un guide simple sur les invites de texte de l'IA : dites à l'IA exactement ce que vous voulez voir

एआई टेक्स्ट प्रॉम्प्ट पर एक सरल गाइड: एआई को ठीक वही बताएं जो आप देखना चाहते हैं

AI テキスト プロンプトに関する簡単なガイド: 見たいものを正確に AI に伝える

AI 텍스트 프롬프트에 대한 간단한 가이드: 보고 싶은 것을 AI에게 정확하게 알려주십시오.

Um guia simples sobre prompts de texto de IA: diga à IA exatamente o que você deseja ver

Yapay Zeka Metin İstemleri Hakkında Basit Bir Kılavuz: Yapay Zekaya Tam Olarak Ne Görmek İstediğinizi Söyleyin

Hướng dẫn đơn giản về lời nhắc văn bản AI: Cho AI biết chính xác những gì bạn muốn xem

AI 文本提示简单指南：准确告诉 AI 你想看什么

Did you know roughly Optimizing your website or webapp for stellar performance is always a crucial goal for any software-based business. half of the users that visit your website leave if it takes more than 3 seconds to load? But, the ecosystem has changed in recent years. Smartphones are taking over. Developers need to build websites and optimize for performance primarily targeting these smaller devices. It’s not solely about performance though. User Experience, called UX for short, is the golden totem few seem to acquire. It has many faces, from the way you turn first-time users that interact with your landing page into paying customers, to the ease-of-use your web app provides paying customers, or how the interface just seems logical. In this tutorial, I’ll explain how the principles of Real User Monitoring, also called End User Monitoring, can help in building and optimizing your single-page application, and how you can easily configure . Let’s go! monitoring a single-page app for free What’s a Good Example of a Single-Page Application? A good single page application has to be performant and blazing fast across all devices. It needs to have great SEO by prerendering dynamic routes. It should be responsive and reliable while serving customers with quick page loads across the whole globe. But, how do you know if your customers are happy, how do you know whether users bounce before even trying your product? You do that by tracking interactions and transactions while monitoring performance and user behavior with solutions. Check out this list of . Real User Monitoring popular tools and solutions Let’s dive into what Real User Monitoring and Single-Page Application Monitoring is, and how you can use it to improve your single-page app. What is Real User Monitoring? Websites and web apps are getting increasingly more complex with more and more client-side logic. This significantly improves UX but adds a lot of code complexity and weight. We’ve come to a point where monitoring only the back end is not enough. To ensure the best possible user experience, you need to monitor the whole client-side as well. The goal of Real User Monitoring is to detect anomalies in real-time when end-user experience is affected by website/webapp performance and unexpected crashes. When your customers are facing poor UX, you need to be notified! Check out for a more detailed walkthrough. this Real User Monitoring guide Having while interacting with your website keeps you on top of insight into how happy your customers are reducing churn and high bounce-rates. With the ability to inspect individual sessions and drill into page-level statistics, you have a birds-eye view of what your customers are facing while interacting with your website or webapp. Tracking page-loads, HTTP requests, page resources, UI interactions, and on-page transactions are all crucial to understanding what your customers are happy and frustrated about. There are tons of Real User Monitoring solutions to choose from, in the next section I’ll walk you through using Sematext Experience. To read more about the tool, check out. It has a nice if you’re starting a side-project and a if you have a business going already. this free tier 30 day free trial of the pro tier How to Monitor Single-Page Applications The sole purpose of Real User Monitoring is to ensure a better experience for the end user. It is not limited to only increasing your web application’s performance, but helping you improve overall satisfaction levels of your users. Single-page application monitoring is crucial for you to monitor virtually anything that executes in the browser. You have insight into performance, page-load times, HTTP response times, and much more. This lets you see what your users are seeing in real-time through . their eyes Why Monitor Single-Page Applications? Having delays in page loads, loading resources, and HTTP requests can all cause your users to become frustrated and ultimately churn, causing you to lose business. That’s why it’s important to track every user’s experience via distinct sessions that provide valuable insight about which pages were accessed, whether they were hard or soft page loads, which HTTP requests were triggered, which device was used, and how the final load time was affected by all these factors. Here’s : a list of important factors and challenges for a single-page application where end-user monitoring can help 1. SEO Single-page apps are inherently terrible regarding SEO because they use JavaScript to load data and content. Meaning the bots that crawl your website and index data for the search engines don’t have any HTML to index because it’s generated once a user loads the SPA in a browser. To remedy this, you can use server-side rendering or prerendering the dynamic SPA routes and generate static HTML. 2. Performance Single-page apps are well known to be fast and responsive websites, because of the way they load content dynamically, without reloading the HTML page. But, you never know what experience your users have while navigating the SPA. If the initial load is way too long or if some routes have bottlenecks with time-consuming HTTP requests. hard 3. Localizing Real User Monitoring gives you a complete overview of the geographical location of your users. You know every detail about page loads and the satisfaction of your users across the whole world. Based on this analysis, localizing your SPA to use languages where you have the most customers is an amazing advantage. 4. Security By tracking users, and their sessions, while interacting with your SPA, you can identify real and fake users, potentially isolating malicious intent. 5. Error and crash reporting Errors happen, your single-page app can crash, it’s all part of life. But, need to know this! Don’t rely on your users telling you. They’ll rather stop using your product than tell you. Real User Monitoring is there to watch your back when JavaScript lets you down. you How is Single-Page Application Monitoring Different from Static Website Monitoring? Monitoring single-page apps and static websites is exactly the same in principle, apart from one important difference. Single-page apps can have . The principle behind how routing works with single-page apps is by using JavaScript to change pages dynamically, without reloading the root HTML page. This is a soft reload and requires one additional configuration watcher dedicated to listening for the routeChange event. Hard reloads are straightforward. They happen when the initial HTML page is loaded for the first time when a user lands on your webapp. hard and soft page loads Sematext Experience covers all the bases, including tracking route changes for single-page apps so you know exactly which pages were loaded hard or soft and how the page rendering behaves. It supports every major single-page app framework or library like Angular, Vue.js, Ember.js, and React, but also a default configuration setting for any framework that uses or . HTML5 pushState Hashchange The way you add Real User Monitoring to your webapp is super simple. Adding Real User Monitoring to a Single-Page Application Here is a list of steps to add Real User Monitoring to a single-page application: First, . Don’t forget to enable the toggle. 1. create an Experience App in Sematext Website uses Single Page Architecture Then, 2. add the Experience Script in your root HTML file’s <head> < = > script type "text/javascript" ( { c=[];e[a]= {c.push( )};
  e[a].queue=c; s=t.createElement(r);s.async= ;s.src=n; u=t.getElementsByTagName(r)[ ];u.parentNode.insertBefore(s,u)})
  ( , , , , ); ( ) function e,t,r,n,a var ( ) function arguments var 1 var 0 window document "script" "//cdn.sematext.com/rum.js" "strum" </ > script < = > script type "text/javascript" strum( , {
    token: ,
    receiverUrl: }); 'config' 'YOUR_TOKEN' 'https://rum-receiver.sematext.com' </ > script Lastly, . Based on what single-page application framework or library you’re using there are specific instructions. Choose what applies to you. 3. add the routeChange event listener to your single-page app’s configuration React React ; { createBrowserHistory createHistory } ; history = createHistory();
history.listen( { (action !== ) {
    strum( , .location.href);
  }
}) { ( );
} import from 'react' import as from 'history' const ( ) => location, action if 'REPLACE' 'routeChange' window export default ( ) function App return ... < = > Router history {history} </ > Router Angular 2+ { Component, OnInit } ; { Router, NavigationStart } ; ({ selector: , templateUrl: }) AppComponent OnInit { ( ) {}
  ngOnInit() { .router.events.subscribe( { (event NavigationStart) {
        strum( , event.url);
      }
    });
  }
} import from '@angular/core' import from '@angular/router' @Component 'app' 'app.component.html' export class implements constructor router: Router private this => event if instanceof 'routeChange' Angular.js $scope.$on( , {
  strum( , .location.href);
}); '$routeChangeStart' ( ) function 'routeChange' window Vue.js < > template < = > div id "app" < /> router-view </ > div </ > template < > script { : , : {
      $route() {
        strum( , .location.href);
      }
    }
  } export default name 'app' watch 'routeChange' document </ > script Ember.js EmberRouter ; { on } ;​

EmberRouter.reopen({ : on( , {
    strum( , .location.href);
  }),
}); Router; import from '@ember/routing/router' import from '@ember/object/evented' doInformAboutRouteChange 'didTransition' ( ) function 'routeChange' window export default Using On-Page Transactions to Track User Experience With the configuration above, you’ll have everything set up and ready to go, apart from custom . These are awesome to have if you want to measure custom events that are important to you in understanding your customers’ behavior. Maybe you want to track transactions when your users go through the sign-up flow or how well they’re onboarding. You can do whatever you want with on-page transactions. Let your imagination go wild! on-page transactions Visualizing User Experience After finishing the configuration, and have data flowing in, you can visualize metrics and start making business decisions to improve the user experience and thus the happiness of your customers. Apart from the obvious, you can also visualize user satisfaction based on which part of the world they live in. Geographical data is crucial to understand and optimize performance for markets you wish to enter. Wrapping Up Monitoring Single-Page Applications Give your customers the experience they deserve. Don’t let performance issues cause churns. Stop relying on your customers to tell you what’s wrong! Diagnose issues right away when they happen. Convert more sales and enhance your single-page application’s performance. Using a Real User Monitoring tool like Sematext Experience gives you a clear overview of who is using your web application, with data from real users and their sessions while interacting with your product. Use this to your advantage, find the hidden problems they encounter to keep user satisfaction high. The state of the Internet today demands lightning-quick web apps that load just as fast on mobile as on desktop. Retain customers who stumble upon your product while casually browsing the web on their phones. Make them want to come back and sign up. Returning to the question from the beginning of this article, Well, now you know. Don’t let yourself be one of those who lose customers before they even land on your website! did you know half of the users that visit your website leave if it takes more than 3 seconds to load? Do you know how fast your website or webapp loads in different browsers and regions, on different devices? With a Real User Monitoring solution, you’ll know. You can check out Sematext , or just shoot me a message on if you want to chat. here Twitter Hope you guys and girls enjoyed reading this as much as I enjoyed writing it. If you liked it, feel free to share so more people see this tutorial. Until next time, be curious and have fun.

SPA Tracking & Monitoring: How to Build Better Single-Page Applications With Real User Monitoring

That’s a short warning to all Tensorflow users working with visual content. Short notice: don’t use any functions! tf.image.resize I was rewriting codebase of our — Let’s Enhance to make it ready for bigger and faster models and API we are working on. As we work with image generation (superresolution, deblurring, etc) we do rely on a typical image processing libraries like OpenCV and PIL. I always had suspicions that it makes sense to use Tensorflow image processing capabilities — in theory, they should be faster. So, I decided to stick to native Tensorflow image preprosessing and dataset building tools using dataset.map() to keep all operations in tensors all around my code. neural network image upscaling service The problem was quite awful — my new and shiny code for superresolution wasn’t able to reproduce not only any state-of-art networks but even my own code that I wrote 4 months ago. And the ugliest part was that results of superresolution itself were pretty good sometimes, the network was working, although not reaching target PSNR and having strange visual artifacts sometimes, like doubling of small lines. Let the debug begin What initially was looking as minor bug became 60 days of struggle and sleepless nights. My faulty logic was simple — there’s something wrong with network definition or training process. Data preprocessing is definitely fine, as I am having meaningful results and visual control over image processing in Tensorboard. I tweaked everything I was able to find, defined network using Keras, Slim, raw TF — nothing, looked for changes in TF 1.3->1.4->1.5 and different CUDA versions, paddings behaviors. I am ashamed even to tell you about my latest suspicions, which involved defects in GPU RAM and statics. I was tweaking perceptual losses and style losses looking for a reason. And each iteration took days to retrain until some meaningful result… Yesterday I found when looking into Tensorboard. It was almost subliminal feeling that something is wrong with the image. I disregarded network output and just overlayed target image and input image(that is a downscaled target image) in Photoshop. Here’s what I got. The Bug, Looks strange, somekind of displacement happening here. Totally againts any logic, this just can’t be true! My code is dead simple. Read image, crop image, resize image. All in Tensorflow. Anyway, RTFM. tf.image.resize_bicubic has a parameter — “align corners”. How in the world would you like to downscale image and not have corners aligned? You can! So there is a very weird behavior of this function known for a long time — . They can’t fix it as this can break lots of old code and pre-trained networks. read this thread Our function isn't even reflection equivariant. It would be lovely to fix this, but I'd be worried about breaking old models. tf.image.resize_area This code does actually displace your image by one pixel to the left and top. Thread suggests that even the interpolation is broken in TensorFlow. It’s 2018, people. Here are the actual downscaling results with TF. Stick to Scipy/OpenCV/numpy/PIL whatever you prefer for image processing. The second I changed it my network worked like a charm (next day actually, when I saw training results).

Stories, News & Mentions about LETSENHANCE

Letsenhance’s Stories on HackerNoon