Here’s an interesting video from Two Minute Papers on OpenAI Five’s recent victory.
Two Minute Papers explores the paper “Learning Correspondence from the Cycle-Consistency of Time” in this video.
Two Minute Papers explores the project that takes NLP to the next level: GPT-2. It’s almost too good – OpenAI has, accordingly, will not release it.
Two Minute Papers explores the paper “ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness ” The source documents and code are available at the following links:
Two Minute Papers has a look at the paper “3D Human Pose Machines with Self-supervised Learning” in his latest video. This topic actually came up in a meeting on Friday.