This week in AI & Machine Learning: Multiple sclerosis gait monitoring, Levi’s 168 years of data, Apple is buying the most AI companies, PyTorch 1.8.1 neural network profiler, Text Classification using GPT2 and more!
Author’s Note:
Our Sense Data Annotation Tool now has more AI-powered features. Learn how features like smart polygon selection, tracking labels forward, and auto labeling can help you save time when preparing data for your computer vision projects here. Or, Get started for free.
🤖 Artificial Intelligence News:
- Machine learning helps spot gait problems in individuals with multiple sclerosis
- Levi’s Katia Walsh on AI: You can do a lot with 168 years of data ← Great example of how you don’t need to be a new company to start using machine learning and how you might already have an advantage with years of data.
- Oil Industry Turns To AI To Help Confront Daunting Challenges ← We’re no strangers when it comes to working with computer vision in the oil and gas industry! Check out some of our oil and gas use cases.
- How AI Can Help Companies Set Prices More Ethically
- Apple bought more AI companies than anyone else between 2016 and 2020
🛠️ Developer Tools & Education:
- Introducing PyTorch Profiler — the new and improved performance tool ← This new deep learning performance profiler in PyTorch 1.8.1 looks awesome!
- Recursive Classification: Replacing Rewards with Examples in Reinforcement Learning
- Constructing Transformers For Longer Sequences with Sparse Attention Methods
- OpenCV Template Matching (cv2.matchTemplate)
- How to Manually Optimize Machine Learning Model Hyperparameters
- CLIP, a model that smokes everyone else
📅 Upcoming Online AI & Data Science Events:
- Brain Meets | Virtual Happy Hour (Apr 7–7:00pm PST)
- Text Classification using GPT2 and Pytorch (Apr 9 9:30am PST)
- PyTorch Ecosystem Day (Apr 21–8:00am PST)
- The Applied ML Summit (Apr 21–9:00am PST) “This is a two-day, free virtual interactive event on data engineering for applied machine learning.”
- Sixgill events ← I’ll be posting computer vision events for April next week, so keep an eye out!.
🎤 Interesting Podcasts & Interviews:
- Recommender systems and high-frequency trading | Practical AI
- Legal Tech, Powered by Machine Learning | SuperDataScience
- Can Language Models Be Too Big? with Emily Bender and Margaret Mitchell | TWiML
- Applying RL to Real-World Robotics with Abhishek Gupta | TWiML ← Applying reinforcement learning in the real world is hard. Learn about some of the common problems and how to overcome them.
- AI+X: AI Innovation in Healthcare | DeepLearningAI
- 3 Questions: Artificial intelligence for health care equity | MIT
📄 Notable Research Papers:
- High-Fidelity Pluralistic Image Completion with Transformers
- USB: Universal-Scale Object Detection Benchmark
- Scaling-up Disentanglement for Image Translation
- Faraway-Frustum: Dealing with Lidar Sparsity for 3D Object Detection using Fusion
- Self-Supervised Training Enhances Online Continual Learning
- Layer-Wise Data-Free CNN Compression
- More Photos are All You Need: Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval
🤝 Connect with AI practitioners of all levels
- Stay connected with artificial intelligence and machine learning practitioners around the world! Slack Group | LinkedIn Group
🦈 About the Author & Sixgill:
- Sage Elliott is a Developer Evangelist at Sixgill & passionate about making AI more approachable. Connect with Sage on Twitter or LinkedIn.
- Sixgill, LLC provides custom enterprise AI solutions, end-to-end machine learning lifecycle management, and fast data annotation for computer vision.