AI Paper Atlas

Research map / 3D vision / Human pose estimation

UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-Aware Curriculum and Iterative Generalist-Specialist Learning

Weikang Wan, Haoran Geng, Yun Liu, Zikang Shan, Yaodong Yang, Li Yi, He Wang

ICCV 2023 · 3D vision › Human pose estimation · approach: Category-level pose estimation · 109 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 category-based object curriculum with geometry-aware curriculum (GeoCurriculum) (ablation)state-based policy · UniDexGrasp trainsuccess rate (%): 79.4 -> 82.7Improved
add curriculum learning versus none (ablation)state-based policy · UniDexGrasp trainsuccess rate (%): 30.5 -> 79.4 (object curriculum) and 82.7 (GeoCurriculum)Improved
use 5 curriculum stages instead of 4 (ablation)GeoCurriculum · UniDexGrasp trainsuccess rate (%): 82.7 -> 82.9No clear effect
add iterative fine-tuning in state-based setting (ablation)state-based policy · UniDexGrasp trainsuccess rate (%): 82.7 -> 84.0Improved
add geometry-aware clustering for specialist assignment in state-based setting (ablation)state-based policy · UniDexGrasp test unseen objectssuccess rate (%): 77.9 -> 84.3Improved
add geometry-aware clustering in vision-based stage (ablation)vision-based policy · UniDexGrasp trainsuccess rate (%): 78.6 -> 85.4Improved
distill state specialists directly to vision generalist instead of via state generalist (ablation)vision-based policy · UniDexGrasp trainsuccess rate (%): 82.7 -> 85.4Improved
add iterative fine-tuning in vision stage (ablation)vision-based policy · UniDexGrasp trainsuccess rate (%): 82.5 -> 85.4Improved
use DAgger-based distillation instead of BC, GAIL or DAPG (ablation)iGSL distillation · UniDexGrasp trainsuccess rate (%): 61.4 (DAPG) -> 87.9Improved
increase number of clusters from 20 to 50 (ablation)GeoClustering · vision-based trainsuccess rate (%): 85.4 -> 77.2Hurt
use geometry-aware clustering instead of category-label clustering (ablation)GeoClustering · vision-based trainsuccess rate (%): 79.7 -> 85.4Improved
use PointNet+Transformer backbone instead of PointNet (ablation)vision-based policy · UniDexGrasp trainsuccess rate (%): 82.1 -> 85.4Improved

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