AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Diffusion Reward: Learning Rewards via Conditional Video Diffusion

Tao Huang, Guangqi Jiang, Yanjie Ze, Huazhe Xu

ECCV 2024 · Reinforcement learning › Visual and representation RL · approach: Goal-conditioned policy learning · 30 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
replace VideoGPT with VQ-Diffusion as video modelvideo prediction for rewards · Real dexterous robot videosPSNR: 23.35 -> 27.46Improved
use conditional entropy reward from video diffusion instead of VIPER log-likelihoodDiffusion Reward · MetaWorld (7 tasks) and Adroit (3 tasks)success rate: +38% (MetaWorld), +35% (Adroit) over best baselineImproved
standardize VIPER log-likelihood reward (VIPER-std)VIPER · MetaWorld and Adroitsuccess rateImproved
use conditional entropy rather than log-likelihood as reward, with diffusion or transformer video models (ablation)Diffusion Reward · MetaWorld (7 tasks)success rateImproved
set diffusion sampling noise to 0 (deterministic reward) (ablation)Diffusion Reward · MetaWorld and Adroitsuccess rateHurt
vary denoising steps from 2 to 20 (ablation)Diffusion Reward · Visual RL taskssuccess rateMixed
vary reward coefficient alpha from 1 to 0 (0.95 best) (ablation)Diffusion Reward · Visual RL taskssuccess rateMixed
increase context length from 1-2 to 4 or 8 historical frames (ablation)Diffusion Reward · Visual RL taskssuccess rateHurt
remove sparse environmental reward from composite reward (ablation)Diffusion Reward · Door and Hammer (Adroit)success rateHurt
increase number of denoised samples M from 1 to 4 (ablation)Diffusion Reward · Door (Adroit)FPS: 86.4 -> 45.9Hurt
use two independent coefficients instead of a single alpha for entropy and RND rewards (ablation)Diffusion Reward · Door and Hammer (Adroit)success rateNo clear effect
apply pretrained reward directly to unseen MetaWorld tasksDiffusion Reward · 5 unseen MetaWorld taskssuccess rateImproved

Later papers that compared with this method

1 later accepted paper reports Diffusion Reward as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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