Differentiable Physics Simulation of Dynamics-Augmented Neural Objects.
Simon Le Cleac'hHong-Xing YuMichelle GuoTaylor A. HowellRuohan GaoJiajun WuZachary ManchesterMac SchwagerPublished in: CoRR (2022)
Keyphrases
- simulation model
- artificial intelligence
- network architecture
- neural network
- data objects
- nonlinear dynamics
- simulation models
- computer science
- d objects
- recurrent networks
- mathematical model
- spatial relationships
- individual objects
- multiple objects
- spiking neurons
- physical processes
- simulation environment
- object model
- object segmentation
- loss function
- objective function
- reinforcement learning