AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Curious Representation Learning for Embodied Intelligence

Yilun Du, Chuang Gan, Phillip Isola

ICCV 2021 · Reinforcement learning › Visual and representation RL · approach: Learned state and action representations · 35 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
freeze pretrained visual weights during downstream RL (ablation)CRL representation · ObjectNav (Matterport3D)Soft SPLImproved
freeze convolutional weights of the from-scratch RL policy (ablation)From Scratch policy · ImageNav/ObjectNavSoft SPL: a fall of over 0.07 SoftSPLHurt
replace random exploration with CRL curiosity policy for gathering contrastive data (ablation)contrastive learning · ImageNav (Gibson)SPL: 0.0285 -> 0.0324Improved
replace random exploration with CRL curiosity policy for gathering contrastive data (ablation)contrastive learning · ObjectNav (Matterport3D)SPL: 0.0042 -> 0.0144Improved
replace learned-counts exploration with CRL (ablation)contrastive learning · ObjectNav (Matterport3D)Soft SPL: 0.110 -> 0.119Improved
replace RND representation with contrastive CRLrepresentation learning · ImageNav (Gibson)Soft SPL: 0.124 -> 0.219Improved
use CRL representation instead of ImageNet initializationfrozen encoder · ImageNav (Gibson)SPL: 0.0193 -> 0.0324Improved
use CRL representation instead of PointNav policy encoderfrozen encoder · ObjectNav (Matterport3D)SPL: 0.0390 -> 0.0144Hurt
freeze convolution weights of a random network (ablation)PointNav agent · PointNavSPLNo clear effect
use CRL representation instead of From ScratchSeq2Seq behavioral cloning · VLN val-unseenSPL: 0.138 -> 0.157Improved
use CRL representation instead of ImageNetSeq2Seq Dagger · VLN val-unseenSPL: 0.206 -> 0.199Hurt
use CRL exploration instead of random exploration (ablation)contrastive model · image diversityLPIPS diversity: 0.708 -> 0.728Improved

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