Papers of the day   All papers

Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video

Comments

Denny Britz: Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video: https://arxiv.org/abs/2005.11212 This is pretty cool. They use a bunch of tricks (first time I hear knot theory) to force an autoencoder representation to be simple enough to extract equations of motion from video https://t.co/g7Ujbt83ip

3 replies, 233 likes


Max Tegmark: We just posted a new #AI paper on how to auto-discover laws of physics from raw warped video with machine learning. It took Silviu & me a year to get this working, using ideas inspired by #generalrelativity & high-dimensional #knottheory - phew! https://arxiv.org/abs/2005.11212 https://t.co/8fW293jq4f

10 replies, 181 likes


arXiv in review: #NeurIPS2020 Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video. (arXiv:2005.11212v1 [cs\.CV]) http://arxiv.org/abs/2005.11212

0 replies, 1 likes


arXiv CS-CV: Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video http://arxiv.org/abs/2005.11212

0 replies, 1 likes


Content

Found on Jun 26 2020 at https://arxiv.org/pdf/2005.11212.pdf

PDF content of a computer science paper: Symbolic Pregression: Discovering Physical Laws from Raw Distorted Video