AI Paper Atlas

Research map / 3D vision / 3D generation

EgoDex: Learning Dexterous Manipulation from Large-Scale Egocentric Video

Ryan Hoque, Peide Huang, David J. Yoon, Mouli sivapurapu, Jian Zhang

ICLR 2026 · 3D vision › 3D generation · approach: Hand-object interaction generation · 232 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
use encoder-decoder instead of decoder-only architectureDec + BC · EgoDex trajectory prediction, 2saverage distance (m), K=1: 0.045 -> 0.044Improved
use flow matching instead of behavior cloningEncDec + BC · EgoDex trajectory prediction, 2saverage distance (m), K=10: 0.044 -> 0.038Improved
use flow matching instead of behavior cloning at a single sampleEncDec + BC · EgoDex trajectory prediction, 2saverage distance (m), K=1: 0.044 -> 0.051Hurt
increase the number of samples K from 1 to 10EncDec + FM · EgoDex trajectory prediction, 2saverage distance (m): 0.051 -> 0.038Improved
shorten prediction horizon from 2s to 1s (ablation)Dec + BC · EgoDex trajectory predictionaverage distance (m): 0.045 -> 0.031Improved
lengthen prediction horizon from 2s to 3s (ablation)Dec + BC · EgoDex trajectory predictionaverage distance (m): 0.045 -> 0.053Hurt
add visual goal image conditioning (ablation)Dec + BC · EgoDex inverse dynamicsfinal distance (m): 0.062 -> 0.029Improved
increase model size from 200M to 500M parameters (ablation)Dec + BC · EgoDex trajectory predictionaverage distance (m): 0.045 -> 0.045No clear effect
increase training dataset size (ablation)Dec + BC · EgoDex trajectory predictionaverage and final distanceImproved
evaluate on out-of-distribution task Stamp PaperDec + BC · EgoDex OOD tasksaverage distance (m): 0.045 -> 0.099Hurt
lengthen prediction horizon from 1s to 3s (ablation)EncDec + FM · EgoDex trajectory prediction, K=10average distance (m): 0.026 -> 0.047Hurt

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