AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

RoboCLIP: One Demonstration is Enough to Learn Robot Policies

Sumedh Anand Sontakke, Jesse Zhang, Séb Arnold, Karl Pertsch, Erdem Biyik, Dorsa Sadigh, Chelsea Finn, Laurent Itti

NeurIPS 2023 · Reinforcement learning › Imitation learning · approach: Imitation combined with RL · 7 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
finetune the RoboCLIP-pretrained policy with a single demonstration of observations and actionsRoboCLIP · Coffee-Push and Faucet-Open (Metaworld)reward: ~200%Improved
use more than one video demonstration for reward generation (ablation)RoboCLIP · in-domain demonstration setupdownstream rewardHurt
replace S3D video-language model with single-image CLIP using the final frame (ablation)RoboCLIP · language-conditioned pretrainingzero-shot performanceHurt
use a randomly initialized S3D instead of a pretrained one (ablation)RoboCLIP · in-domain video demonstration setuppretraining reward signalHurt
edit a button-pushing video embedding by subtracting 'button' text and adding 'drawer' textRoboCLIP · Drawer-Closezero-shot scoreImproved
use RoboCLIP language reward instead of GAIL and AIRL baselinespretraining reward · Drawer-Close, Door-Close, Button-Presstrue task rewardImproved
use RoboCLIP out-of-domain video reward instead of GAIL and AIRLpretraining reward · Button-Press, Drawer-Open, Door-Openzero-shot task performance: 2-3xImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy