AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

PQDA:Policy-Aligned Q-Consistency Meets Decoupled Augmentation for Generalizable Visual RL

Yun Zhou, Yuqiang Wu, Chunyu Tan

AAAI 2026 · Reinforcement learning › Visual and representation RL · approach: Robust visual encoders for RL

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
add policy-aligned Q-consistency (PAQC) to SAC (ablation)SAC · CVDMC color-easy video-easyaverage return: 548 -> 606Improved
add foreground-background decoupled augmentation (FBDA) on top of PAQC (ablation)SAC w/ PAQC · CVDMC color-easy video-easyaverage return: 606 -> 822Improved
replace original Q-regularization with PAQC (ablation)SMG · CVDMC color-easy video-easyaverage return: 763 -> 791Improved
replace original Q-regularization with PAQC (ablation)SGQN · CVDMC color-easy video-easyaverage return: 691 -> 753Improved
replace PAQC with original Q-consistency (ablation)PQDA · CVDMC color-easy video-easyaverage return: 822 -> 797Hurt
use GMM mask instead of ground-truth mask (ablation)PQDA · CVDMC color-easy video-easyaverage return: 816 -> 822No clear effect
replace SGQN saliency mask with GMM mask (ablation)SGQN · CVDMC color-easy video-easyaverage return and training time: 691 -> 803; 13h -> 10.5hImproved
replace SMG mask with GMM mask (ablation)SMG · CVDMC color-easy video-easyaverage return and training time: 763 -> 808; 17h -> 13.5hImproved
use FBDA instead of background-only augmentation (ablation)PQDA · CVDMC color-easy video-easyaverage return: 772 -> 822Improved
use FBDA instead of random overlay augmentation (ablation)PQDA · CVDMC color-easy video-easyaverage return: 761 -> 822Improved
use PQDA instead of SMGvisual RL generalization · DMControl video-hard, cheetah runreturn: 303±46 (SMG) vs 185±42Hurt
use PQDA instead of SAC and SGQN regarding training costvisual RL · DMControl with color and video distractionsreward and training time: 217 -> 810 reward; 5.5h -> 6.5hImproved

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