This week in AI & Machine Learning: USA launches AI.gov, $100 Million to accelerate gene therapy, let a sommel-AI-er choose your wine, robotics in healthcare, fastai v2.3.1, data annotation for computer vision, and more!
🤖 Artificial Intelligence News:
- The USA Federal Government Launches AI.gov ← If you’ve been following these weekly updates for awhile, you may remember the government publicly ramped up its AI efforts in the past year. Now they have an official website.
- AI is Ready to Take on Massive Healthcare Challenge
- Mayo Clinic trial signals potential for AI-guided heart disease detection
- Vivino: Choose Your Next Great Wine With Big Data And Artificial Intelligence ← Finally everyone can have their own sommelier. I’d personally call it a sommel-AI-er.
- Dyno Therapeutics Closes $100 Million Series A to Accelerate AI-powered Gene Therapy Platform
- How AI Robotics are Transforming the Healthcare Industry ← A neat list of how AI & robotic applications can be used in the healthcare industry!
🛠️ Developer Tools & Education:
- Fastai v2.3.1 released ← this update now contains official Windows support!
- Using TFX inference with Dataflow for large scale ML inference patterns
- Matplotlib 3.4.0 is out ← this release has some cool new features like IPython representations for Colormap objects!
- Building a TinyML Application with TF Micro and SensiML
- pytorch-lightning 1.3.0 release ← lighting now has new early stopping strategies, improved TPU support, and more.
- Face Recognition with Local Binary Patterns (LBPs) and OpenCV
🎤 Interesting Podcasts & Interviews:
- Walmart’s Grant Gelvin on Prediction Analytics at Supercenter Scale
- Generating “hunches” using smart home data ← Smart home data is complicated!
- Dask + Data Science Careers with Jacqueline Nolis
- Building a Unified NLP Framework at LinkedIn with Huiji Gao
- High-Impact Data Science Made Easy
📄 Notable Research Papers:
- Memory-Efficient Semi-Supervised Continual Learning: The World is its Own Replay Buffer
- Deep Learning based Multi-modal Computing with Feature Disentanglement for MRI Image Synthesis
- Object-centric Video Prediction without Annotation
- MLP-Mixer: An all-MLP Architecture for Vision
- How Fine-Tuning Allows for Effective Meta-Learning
- Self-Supervised Multi-Frame Monocular Scene Flow
- 3D Vehicle Detection Using Camera and Low-Resolution LiDAR
- PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation
- Aligning Subtitles in Sign Language Videos