If you&#x27;re seeing this message, that means JavaScript has been disabled on your browser · Please enable JavaScript to make this website work

Tech in Asia - Connecting Asia's startup ecosystem

Fintech startup <strong>BharatX</strong> acquires healthcare finance firm Zenifi to enter medical lending. Zenifi CEO Padam Kataria joins <strong>BharatX</strong> to lead healthcare lending vertical.

BharatX: BharatX expands into healthcare finance with acquisition of Zenifi - The Economic Times

BharatX is a VC Backed early-stage FinTech startup that uses technology to empower the Indian Middle Class with credit. Approximately 110 Million Households whose income is more than 5 lakh per annum, do not get access to Formal Credit. BharatX wants to change that by enabling Credit-as-a-Feature on Consumer facing apps! Their access to data allows them to innovate with products which remove barriers and lowers costs at the same time. They are changing the way credit is made available to the Indian middle class. LinkedIn Cover Attribution: Wikimedia Commons/Praveenpaavni.

BharatX

The Binance Build for Bharat hackathon - by , the global blockchain company behind the world’s largest digital asset exchange and blockchain protocol - Binance Smart Chain, announced Semiott Protocol, and as the top three winners. This was the first DeFi-focused Hackathon in India, based on the vision of the initiative - : an initiative designed under the . The Hackathon also gained the support of large scale tech players: ( Previously Matic Network ), , , . Binance Whrrl GoSats Polaris Accelerator Programme Blockchain for India Fund Google Cloud, Polygon Marlin Protocol Band Protocol Ankr cloud infrastructure The winners were selected from a pool of 350+ developers and startups that had applied for the Build for Bharat programme and were challenged with multiple tasks including a 48 hours long hackathon and Binance Smart Chain integrations. The top 8 finalists battled it out at a virtual Demo Day with participants attending from across the globe. The event also saw an opening address by Changpeng Zhao, CEO, Binance and a panel discussion featuring Nischal Shetty, CEO, WazirX , Ryan Fang, Coo and Co-Founder, Ankr, Pratik Reddy, Head of Research at Marlin Protocol and Raghu Mohan, CEO, Lumos labs. The winner of the Hackathon, the Semiott Protocol by , a serial entrepreneur, is an algorithmic stablecoin for value veracity. , the first runners up, is a blockchain platform connecting custodian of assets (warehouse) and banks with integrated commodity trading platforms. Gokul Alex Whrrl , the second runners-up, is a Bitcoin stacking app that lets users earn bitcoin rewards when they shop with partnered merchants. GoSats The Top 3 winning projects will secure prizes from a pool of 40,000 USD that includes Google Cloud Credits. “This hackathon has been an interesting and enriching experience. We are happy to provide the catalyst for the growth of the DeFi economy in India.” said Changpeng Zhao (CZ), CEO at Binance. “The disruptive projects and ideas that we have seen through this engagement has been a testimony to India’s readiness for a decentralized economy and at Binance we are happy to provide the necessary fuel to augment this innovation.” he added. Build for Bharat aimed to identify creators, innovators, startups, and developers from all over India working on unique and creative products utilizing digital assets and blockchain technology around decentralised finance (DeFi) such as mini-Tokens, micro financing products, blockchain datasets, blockchain-AI solutions and more. Additionally, this month Binance launched the under the $100M fund set up to connect DeFi and CeFi. The program aims to help BSC-based startups gain industry exposure and accelerate their business growth. Participants from existing hackathons, including Build for Bharat can participate in the programme. Applications for the first round of MVB Programme will close by 28th Feb. Most Valuable Builder Programme

Meet the Winners of Binance's Build for Bharat Hackathon 

Episode 17: An interview with Bharath Rao, CEO and founder of CryptoDisrupted http://LEVERJ.io Trent interviews Bharath Rao, CEO of . In the interview we discuss decentralized exchanges, derivatives trading, the 2008 market crash, and the future of crypto trading. http://LEVERJ.io Also . available on iTunes For more episodes of , or . Crypto Disrupted subscribe on YouTube subscribe and listen on iTunes

Decentralized Leveraged Exchange with Bharath Rao

Authors:
(1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore;
(2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore;
(3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore;
(4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore;
(5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore;
(6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction II Method III Datasets IV Experiments, Acknowledgment, and References IV. EXPERIMENTS We test three model configurations on both the simulator and tank datasets. First, we implement a pose regression model based on GoogLeNet [7] as our baseline model. Secondly, we implement a model using ResNet-50 [9] to evaluate the improvement possible by use of a deeper network. ResNet allows us to have deeper networks with more free parameters, and tries to avoid the problem of overfitting by having residual connections between layers [9]. Thirdly, we implement a pose regression model using an LSTM as an intermediate layer. The results indicate that all three configurations can perform well in both simulated and tank datasets as shown in Table I. The errors are minimal and of comparable magnitude to the noise in the pose recorded by the camera sensors. Fig. 4 shows that the trajectories predicted by the models are very close to the actual trajectories in terms of both position and orientation. With the simulator dataset, which is free of noise, turbidity, light distortion, and other challenges that may be observed in real underwater settings, the ResNet-50 and LSTM-based model performed better than the baseline model. On the other hand, with the tank datasets with some distortions typical of underwater scenarios, ResNet-50 performed worse than the baseline model, whereas the LSTM model did better than baseline with dataset #1 (which mostly featured translation but almost no rotation). This suggests the ResNet-50 architecture may be slightly overfitting despite the regularization, so it is not able to provide better performance in the tank dataset. We also note that data augmentation significantly improves the model performance, so in cases where there is no significant rotation, the use of data from both cameras can be used to bolster performance. Overall, these methods are robust for application in real underwater environments, and show promise for use in open water settings, where they will be tested next. ACKNOWLEDGMENT This research project is supported by A*STAR under its RIE2020 Advanced Manufacturing and Engineering (AME) Industry Alignment Fund - Pre-Positioning (IAF-PP) Grant No. A20H8a0241. REFERENCES [1] J. Gonzalez-Garcıa, A. Gomez-Espinosa, E. Cuan-Urquizo, L. G. ´ Garcıa-Valdovinos, T. Salgado-Jimenez, and J. A. E. Cabello, “Au-tonomous Underwater Vehicles: Localization, Navigation, and Commu- nication for Collaborative Missions,” Applied Sciences, vol. 10, no. 4, p. 1256, Feb. 2020. [2] L. Peng and M. Chitre, ”Regressing Poses from Monocular Images in an Underwater Environment”, 2022 OCEANS - Chennai, in press. [3] A. Kendall, M. Grimes and R. Cipolla, ”PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization,” 2015 IEEE International Conference on Computer Vision (ICCV), 2015, pp. 2938- 2946, doi: 10.1109/ICCV.2015.336. [4] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, Nov. 1997, doi: 10.1162/neco.1997.9.8.1735. [5] F. Walch, C. Hazirbas, L. Leal-Taixe, T. Sattler, S. Hilsenbeck, and D. Cremers, “Image-based localization using lstms for structured feature correlation,” 2017 IEEE International Conference on Computer Vision (ICCV), 2017. [6] A. Kendall and R. Cipolla, ”Geometric Loss Functions for Camera Pose Regression with Deep Learning,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 6555-6564, doi: 10.1109/CVPR.2017.694. [7] C. Szegedy et al., ”Going deeper with convolutions,” 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 1-9, doi: 10.1109/CVPR.2015.7298594. [8] A. Chaudhary, R. Mishra, B. Kalyan, and M. Chitre, ”Development of an Underwater Simulator using Unity3D and Robot Operating Systems”, 2021 OCEANS - San Diego, in press. [9] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Authors: Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction I Introduction II Method II Method III Datasets III Datasets IV Experiments, Acknowledgment, and References IV Experiments, Acknowledgment, and References IV. EXPERIMENTS We test three model configurations on both the simulator and tank datasets. First, we implement a pose regression model based on GoogLeNet [7] as our baseline model. Secondly, we implement a model using ResNet-50 [9] to evaluate the improvement possible by use of a deeper network. ResNet allows us to have deeper networks with more free parameters, and tries to avoid the problem of overfitting by having residual connections between layers [9]. Thirdly, we implement a pose regression model using an LSTM as an intermediate layer. The results indicate that all three configurations can perform well in both simulated and tank datasets as shown in Table I. The errors are minimal and of comparable magnitude to the noise in the pose recorded by the camera sensors. Fig. 4 shows that the trajectories predicted by the models are very close to the actual trajectories in terms of both position and orientation. With the simulator dataset, which is free of noise, turbidity, light distortion, and other challenges that may be observed in real underwater settings, the ResNet-50 and LSTM-based model performed better than the baseline model. On the other hand, with the tank datasets with some distortions typical of underwater scenarios, ResNet-50 performed worse than the baseline model, whereas the LSTM model did better than baseline with dataset #1 (which mostly featured translation but almost no rotation). This suggests the ResNet-50 architecture may be slightly overfitting despite the regularization, so it is not able to provide better performance in the tank dataset. We also note that data augmentation significantly improves the model performance, so in cases where there is no significant rotation, the use of data from both cameras can be used to bolster performance. Overall, these methods are robust for application in real underwater environments, and show promise for use in open water settings, where they will be tested next. ACKNOWLEDGMENT This research project is supported by A*STAR under its RIE2020 Advanced Manufacturing and Engineering (AME) Industry Alignment Fund - Pre-Positioning (IAF-PP) Grant No. A20H8a0241. REFERENCES [1] J. Gonzalez-Garcıa, A. Gomez-Espinosa, E. Cuan-Urquizo, L. G. ´ Garcıa-Valdovinos, T. Salgado-Jimenez, and J. A. E. Cabello, “Au-tonomous Underwater Vehicles: Localization, Navigation, and Commu- nication for Collaborative Missions,” Applied Sciences, vol. 10, no. 4, p. 1256, Feb. 2020. [2] L. Peng and M. Chitre, ”Regressing Poses from Monocular Images in an Underwater Environment”, 2022 OCEANS - Chennai, in press. [3] A. Kendall, M. Grimes and R. Cipolla, ”PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization,” 2015 IEEE International Conference on Computer Vision (ICCV), 2015, pp. 2938- 2946, doi: 10.1109/ICCV.2015.336. [4] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, Nov. 1997, doi: 10.1162/neco.1997.9.8.1735. [5] F. Walch, C. Hazirbas, L. Leal-Taixe, T. Sattler, S. Hilsenbeck, and D. Cremers, “Image-based localization using lstms for structured feature correlation,” 2017 IEEE International Conference on Computer Vision (ICCV), 2017. [6] A. Kendall and R. Cipolla, ”Geometric Loss Functions for Camera Pose Regression with Deep Learning,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 6555-6564, doi: 10.1109/CVPR.2017.694. [7] C. Szegedy et al., ”Going deeper with convolutions,” 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015, pp. 1-9, doi: 10.1109/CVPR.2015.7298594. [8] A. Chaudhary, R. Mishra, B. Kalyan, and M. Chitre, ”Development of an Underwater Simulator using Unity3D and Robot Operating Systems”, 2021 OCEANS - San Diego, in press. [9] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Underwater Visual Localization Using Machine Learning and LSTM: Experiments, and References

Authors:
(1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore;
(2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore;
(3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore;
(4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore;
(5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore;
(6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction II Method III Datasets IV Experiments, Acknowledgment, and References III. DATASETS To train and test our model, we used one dataset collected from an underwater robotics simulator [8] as shown in Fig. 3 as well as two datasets collected from a tank as shown in Fig. 2. In the simulator dataset, we operated the ROV in simulation to perform inspection on a vertical pipe in a spiral motion. The total spatial extent covered by the ROV during the inspection is about 2×4×2 m. 14,400 samples of image-pose pair data were collected. In the first tank dataset, we operated the ROV in a lawnmower path with translations only (Fig.2), and the rotations were minimal. We collected 3,437 data samples. In the second tank dataset, the ROV primarily performed rotation maneuvers at 5 selected points. We collected 4,977 data samples. We augmented the left-camera dataset by adding the right-camera data, and thereby using the geometry of the stereo camera placement to provide more training data. This worked well and yielded better performance. The total spatial extent covered in tank datasets was 0.4×0.6×0.2 m. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Authors: Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction I Introduction II Method II Method III Datasets III Datasets IV Experiments, Acknowledgment, and References IV Experiments, Acknowledgment, and References III. DATASETS To train and test our model, we used one dataset collected from an underwater robotics simulator [8] as shown in Fig. 3 as well as two datasets collected from a tank as shown in Fig. 2. In the simulator dataset, we operated the ROV in simulation to perform inspection on a vertical pipe in a spiral motion. The total spatial extent covered by the ROV during the inspection is about 2×4×2 m. 14,400 samples of image-pose pair data were collected. In the first tank dataset, we operated the ROV in a lawnmower path with translations only (Fig.2), and the rotations were minimal. We collected 3,437 data samples. In the second tank dataset, the ROV primarily performed rotation maneuvers at 5 selected points. We collected 4,977 data samples. We augmented the left-camera dataset by adding the right-camera data, and thereby using the geometry of the stereo camera placement to provide more training data. This worked well and yielded better performance. The total spatial extent covered in tank datasets was 0.4×0.6×0.2 m. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Underwater Visual Localization Using Machine Learning and LSTM: Datasets

Authors:
(1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore;
(2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore;
(3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore;
(4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore;
(5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore;
(6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction II Method III Datasets IV Experiments, Acknowledgment, and References II. METHOD We implement two different architectures. The first uses pretrained deep convolutional neural networks (DCNN) to extract features from the input images. The resulting feature map is then passed to an affine regressor consisting of dense neural layers to output a 7-dimensional pose vector estimates. The second architecture adds an additional LSTM layer between the DCNN and affine regressor to perform dimensionality reduction by assessing the spatial structure of the image. This is done by four LSTMs which parse the DCNN-output feature map in different directions starting at each of the four image corners to process the spatial trends, and compress it into lower-dimensional information which is easier processed by the affine regressor. The input images used in the training are rescaled to 256×256 pixels before cropping into a 224×224 feature input using centre cropping. To speed up training, the images are normalized by the mean and standard deviation of the images. The poses are also normalized to lie within [-1, 1]. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Authors: Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction I Introduction II Method II Method III Datasets III Datasets IV Experiments, Acknowledgment, and References IV Experiments, Acknowledgment, and References II. METHOD We implement two different architectures. The first uses pretrained deep convolutional neural networks (DCNN) to extract features from the input images. The resulting feature map is then passed to an affine regressor consisting of dense neural layers to output a 7-dimensional pose vector estimates. The second architecture adds an additional LSTM layer between the DCNN and affine regressor to perform dimensionality reduction by assessing the spatial structure of the image. This is done by four LSTMs which parse the DCNN-output feature map in different directions starting at each of the four image corners to process the spatial trends, and compress it into lower-dimensional information which is easier processed by the affine regressor. The input images used in the training are rescaled to 256×256 pixels before cropping into a 224×224 feature input using centre cropping. To speed up training, the images are normalized by the mean and standard deviation of the images. The poses are also normalized to lie within [-1, 1]. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Underwater Visual Localization Using Machine Learning and LSTM: Method

Authors:
(1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore;
(2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore;
(3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore;
(4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore;
(5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore;
(6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction II Method III Datasets IV Experiments, Acknowledgment, and References I. INTRODUCTION Visual localization is a potential solution for the problem of localization in a known underwater environment for inspection. Inspection missions may often involve operations around marine structures, making acoustic navigation with beacons difficult due to shadowing and multipath [1]. As the vehicle has to operate close to structures, inertial navigation systems, which accumulate errors with time, may not provide sufficient positioning accuracy [1]. In comparison, visual localization using cameras may offer a cost-effective, consistent and accurate alternative in such missions. Previous work has shown that machine learning-based regression methods based on PoseNet [3], can effectively regress a 6-degree-offreedom (DOF) pose from a single 224×224 RGB image with approximately 6 cm position accuracy and 1.7°orientation accuracy when tested on simulated underwater datasets [2]. It was also shown that using a deeper neural network as the extractor may improve the model’s localization accuracy [2]. This work further investigates the effectiveness of such models on underwater datasets and also explores different techniques to further improve localization performance. We have three main contributions: We explore the use of long-short-term memory (LSTM) [4] in the pose regression model to exploit spatial correlation of the image features and to achieve more structured dimensionality reduction [5].


We test the proposed models on underwater datasets collected from a 1.6 m × 1 m × 1 m water-filled tank using a remotely operated vehicle (ROV). The tank offers an environment where we can control lighting and turbidity. The models are able to achieve good accuracy in these datasets, with performance comparable to that obtained with the simulator dataset.


The base dataset consist of images taken from the first camera of a stereo camera mounted on the vehicle. Furthermore, we explore the performance improvement obtained by augmenting the data with additional images from the second camera. Fig. 2 shows some examples of underwater scenes in the tank dataset. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Authors: Authors: (1) Luyuan Peng, Acoustic Research Laboratory, National University of Singapore; (2) Hari Vishnu, Acoustic Research Laboratory, National University of Singapore; (3) Mandar Chitre, Acoustic Research Laboratory, National University of Singapore; (4) Yuen Min Too, Acoustic Research Laboratory, National University of Singapore; (5) Bharath Kalyan, Acoustic Research Laboratory, National University of Singapore; (6) Rajat Mishra, Acoustic Research Laboratory, National University of Singapore. Table of Links I Introduction I Introduction II Method II Method III Datasets III Datasets IV Experiments, Acknowledgment, and References IV Experiments, Acknowledgment, and References I. INTRODUCTION Visual localization is a potential solution for the problem of localization in a known underwater environment for inspection. Inspection missions may often involve operations around marine structures, making acoustic navigation with beacons difficult due to shadowing and multipath [1]. As the vehicle has to operate close to structures, inertial navigation systems, which accumulate errors with time, may not provide sufficient positioning accuracy [1]. In comparison, visual localization using cameras may offer a cost-effective, consistent and accurate alternative in such missions. Previous work has shown that machine learning-based regression methods based on PoseNet [3], can effectively regress a 6-degree-offreedom (DOF) pose from a single 224×224 RGB image with approximately 6 cm position accuracy and 1.7°orientation accuracy when tested on simulated underwater datasets [2]. It was also shown that using a deeper neural network as the extractor may improve the model’s localization accuracy [2]. This work further investigates the effectiveness of such models on underwater datasets and also explores different techniques to further improve localization performance. We have three main contributions: We explore the use of long-short-term memory (LSTM) [4] in the pose regression model to exploit spatial correlation of the image features and to achieve more structured dimensionality reduction [5]. We test the proposed models on underwater datasets collected from a 1.6 m × 1 m × 1 m water-filled tank using a remotely operated vehicle (ROV). The tank offers an environment where we can control lighting and turbidity. The models are able to achieve good accuracy in these datasets, with performance comparable to that obtained with the simulator dataset. The base dataset consist of images taken from the first camera of a stereo camera mounted on the vehicle. Furthermore, we explore the performance improvement obtained by augmenting the data with additional images from the second camera. Fig. 2 shows some examples of underwater scenes in the tank dataset. We explore the use of long-short-term memory (LSTM) [4] in the pose regression model to exploit spatial correlation of the image features and to achieve more structured dimensionality reduction [5]. We explore the use of long-short-term memory (LSTM) [4] in the pose regression model to exploit spatial correlation of the image features and to achieve more structured dimensionality reduction [5]. We test the proposed models on underwater datasets collected from a 1.6 m × 1 m × 1 m water-filled tank using a remotely operated vehicle (ROV). The tank offers an environment where we can control lighting and turbidity. The models are able to achieve good accuracy in these datasets, with performance comparable to that obtained with the simulator dataset. We test the proposed models on underwater datasets collected from a 1.6 m × 1 m × 1 m water-filled tank using a remotely operated vehicle (ROV). The tank offers an environment where we can control lighting and turbidity. The models are able to achieve good accuracy in these datasets, with performance comparable to that obtained with the simulator dataset. The base dataset consist of images taken from the first camera of a stereo camera mounted on the vehicle. Furthermore, we explore the performance improvement obtained by augmenting the data with additional images from the second camera. Fig. 2 shows some examples of underwater scenes in the tank dataset. The base dataset consist of images taken from the first camera of a stereo camera mounted on the vehicle. Furthermore, we explore the performance improvement obtained by augmenting the data with additional images from the second camera. Fig. 2 shows some examples of underwater scenes in the tank dataset. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

BHARATX

bharatx.tech

11-50 employees

Since 2019

COMPANY RANKING

BHARATX

EVERGREEN INDEX #2207

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

BharatX WIKI

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

BHARATX

bharatx.tech

11-50 employees

Since 2019

COMPANY RANKING

BHARATX

EVERGREEN INDEX #2207

RELATED COMPANIESRANK

STORIES ABOUT BHARATX

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

BharatX WIKI

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