AI Paper Atlas

Research map / 3D vision / Human pose estimation

PartManip: Learning Cross-Category Generalizable Part Manipulation Policy From Point Cloud Observations

Haoran Geng, Ziming Li, Yiran Geng, Jiayi Chen, Hao Dong, He Wang

CVPR 2023 · 3D vision › Human pose estimation · approach: Category-level pose estimation · 29 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
Add part-pose canonicalization to the state-based expert (ablation)PartManip state-based expert · OpenDoor, Training Setsuccess rate (%): 67.8±3.4 -> 82.2±0.2Improved
Add part-pose canonicalization to the state-based expert (ablation)PartManip state-based expert · OpenDoor, Val-Inter Setsuccess rate (%): 23.4±3.9 -> 50.7±4.1Improved
Use DAgger instead of directly training with PPO (ablation)PartManip vision policy · OpenDoor, Val-Inter Setsuccess rate (%): 0.2±0.2 -> 31.5±2.9Improved
Add point-cloud augmentation and Sparse-UNet to DAgger (ablation)PartManip vision policy · OpenDoor, Val-Inter Setsuccess rate (%): 31.5±2.9 -> 41.7±2.5Improved
Add expert-demonstration pretraining (ablation)PartManip vision policy with DAgger, augmentation, Sparse-UNet · OpenDoor, Val-Inter Setsuccess rate (%): 41.7±2.5 -> 37.2±1.2Hurt
Add domain-adversarial learning (ablation)PartManip vision policy with DAgger, augmentation, Sparse-UNet, pretraining · OpenDoor, Val-Inter Setsuccess rate (%): 37.2±1.2 -> 49.1±1.5Improved
Add domain-adversarial learning (ablation)PartManip vision policy with DAgger, augmentation, Sparse-UNet, pretraining · OpenDrawer, Val-Inter Setsuccess rate (%): 48.8±4.5 -> 54.7±4.2Improved
Use PartManip rather than the strongest listed baselineOpenDoor benchmark · OpenDoor Val-Inter Setsuccess rate (%): 26.9±2.2 -> 49.1±1.5Improved
Use PartManip rather than the strongest listed baselineOpenDrawer benchmark · OpenDrawer Val-Inter Setsuccess rate (%): 31.2±5.1 -> 54.7±4.2Improved

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