IMLE Policy: Fast and Sample Efficient Visuomotor Policy Learning via Implicit Maximum Likelihood Estimation- state of art BC comparing to diffusion- faster inference speed#RSS0225 #BCX Overview, Project Website

IMLE Policy: Fast and Sample Efficient Visuomotor Policy Learning via Implicit Maximum Likelihood Estimation

- state of art BC comparing to diffusion
- faster inference speed

#RSS0225 #BC

X Overview, Project Website

Are Diffusion and Flow Matching the best generative modelling algorithms for behaviour cloning in robotics?

✅Multimodality
❌Fast, Single-Step Inference
❌Sample Efficient

💡 We introduce IMLE Policy, a novel behaviour cloning approach that can satisfy all…