Skip to content

Slip or Not? Unsupervised Learning to Understand Physical Scene Using Multimodal Variational Physics Inference Network

Notifications You must be signed in to change notification settings

frankaging/generative-physics-inference

Repository files navigation

generative-physics-inference

Slip or Not? Unsupervised Learning to Understand Physical Scene Using Multimodal Variational Physics Inference Network Zhengxuan Wu; Xiyu Zhang Stanford University

install

pytorch

run

python trainMVAE.py --visualize

dataset

You can generative datasets using this simulation engine. https://github.com/frankaging/2d-static-friction-simulation

multimodal

This model is based on MVAE. It could be used to do multimodal physics scene understandings.

About

Slip or Not? Unsupervised Learning to Understand Physical Scene Using Multimodal Variational Physics Inference Network

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages