This week in AI & Machine Learning: GPT-3 in Azure, smart cameras, fully automated farms, TensorFlow Decision trees, handwriting recognition, and more!
🤖 Artificial Intelligence News:
- From conversation to code: Microsoft introduces its first product features powered by GPT-3 ← We finally get to see a peak of what Microsoft is doing with the exclusive GPT-3 language model license!
- Tesla aspires to become a robotics firm in Artificial Intelligence
- Robots and artificial intelligence to guide Australia’s first fully automated farm “Robots and artificial intelligence will replace workers on Australia’s first fully automated farm created at a cost of $20 million.”
- Synthetic data’s growing role in healthcare AI, machine learning and robotics
🛠️ Developer Tools & Education:
- Introducing TensorFlow Decision Forests “We are happy to open source TensorFlow Decision Forests (TF-DF). TF-DF is a collection of production-ready state-of-the-art algorithms for training, serving and interpreting decision forest models (including random forests and gradient boosted trees).”
- Dynaboard: Moving beyond accuracy to holistic model evaluation in NLP
- Understanding Contextual Facial Expressions Across the Globe
- Grid search hyperparameter tuning with scikit-learn ( GridSearchCV )
- Everything you need to know about TorchVision’s MobileNetV3 implementation
🎤 Interesting Podcasts & Interviews:
- Simon Knowles on pushing AI computing to the limit by rethinking chips
- Diving Deep into Synthetic Data with Alex Watson of Gretel.ai
- Making artificial intelligence practical, productive, and accessible to everyone
- Machine Learning at NVIDIA
- Applied AI Research at AWS with Alex Smola
- Causal Models in Practice at Lyft with Sean Taylor
📄 Notable Research Papers:
- SiamMOT: Siamese Multi-Object Tracking
- ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reduction
- AndroidEnv: A Reinforcement Learning Platform for Android
- True Few-Shot Learning with Language Models
- Synthetic Data Generation for Grammatical Error Correction with Tagged Corruption Models
