This week in AI & Machine Learning: Lyft sells self-driving car division, robotaxis, AI’s carbon footprint, robot makes paella, self-supervised learning with transformers, RNA Sequencing, how drones are reshaping Africa, and more!
🤖 Artificial Intelligence News:
- Toyota snaps up Lyft’s self-driving cars unit for $550 million ← Similar to the outcome as the Uber self-driving team. It’s interesting to see both major ride sharing services depart from their once bullish stance on building self-driving cars.
- Keeping Employees & Workplaces Safe with Computer Vision AI
- Baidu Will Launch Driverless Robotaxis In Beijing This Week ← Exciting to see another company put autonomous vehicles on roads!
- Google-led paper pushes back against claims of AI inefficiency ← Google shows research suggesting large AI models don’t have to have a high carbon footprint.
- Artificial intelligence is learning how to dodge space junk in orbit
- Computer vision in AI: The data needed to succeed
- Robot that can cook a paella is causing quite a stir in Spain
🛠️ Developer Tools & Education:
- Advancing the state of the art in computer vision with self-supervised Transformers and 10x more efficient training ← Facebook AI is combining self-supervised learning and transformers to deliver some awesome results!
- Dash Apps that Analyze RNA Sequencing Data for Drug & Vaccine Development
- Fastdebug ← A helpful library for improving torch and fastai errors.
- Adaptive Framework for On-device Recommendation “We are excited to announce an adaptive framework to build on-device recommendation ML solutions with your own data and advanced user modeling architecture.”
- PyTorch implementation of GLOM “Geoffrey Hinton’s new idea that integrates concepts from neural fields, top-down-bottom-up processing, and attention (consensus between columns).”
- Model-Based RL for Decentralized Multi-agent Navigation
- scikit-learn 0.24.2 is out ← Mostly just minor fixes.
🎤 Interesting Podcasts & Interviews:
- Keenan Wyrobek discusses how Zipline’s AI-drones are reshaping Africa
- Mapping the world with CARMERA and “road intelligence”
- AI Storytelling Systems with Mark Riedl
- Machine Learning for Equitable Healthcare Outcomes with Irene Chen
📄 Notable Research Papers:
- BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignment
- Explaining in Style: Training a GAN to explain a classifier in StyleSpace
- Unsupervised 3D Shape Completion through GAN Inversion
- MVS2D: Efficient Multi-view Stereo via Attention-Driven 2D Convolutions
- Every Annotation Counts: Multi-label Deep Supervision for Medical Image Segmentation