Research map / 3D vision / 3D generation
NL2Contact: Natural Language Guided 3D Hand-Object Contact Modeling with Diffusion Model
Zhongqun Zhang, Hengfei Wang, Ziwei Yu, Yihua Cheng, Angela Yao, Hyung Jin Chang
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add natural language contact description guidance (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 24.3 -> 21.7 | Improved |
| use more specific low-level text instead of high-level text (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 23.9 -> 21.7 | Improved |
| add cascaded text-to-hand attention fusion (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 23.5 -> 21.7 | Improved |
| add text-to-hand-object fusion into point features (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 22.9 -> 21.7 | Improved |
| use staged diffusion (hand pose first, then contact) rather than a single stage (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 27.3 -> 21.7 | Improved |
| use diffusion instead of VAE generators (ablation) | NL2Contact · ContactDescribe | MPJPE (mm): 21.9 -> 21.7 | Improved |
| use NL2Contact contact maps vs ContactOpt for grasp optimization | grasp pose optimization · ContactDescribe | intersection volume: 12.8 -> 7.1 | Improved |
| use NL2Contact vs S2Contact on Described HO3D | grasp pose optimization · Described HO3D | intersection volume (mm^3): 3.52 vs 4.39 | Hurt |
| use NL2Contact vs S2Contact on Described HO3D | grasp pose optimization · Described HO3D | MPJPE (mm): 8.7 -> 8.4 | Improved |
| use NL2Contact vs ContactGen for grasp generation | human grasp generation · unseen HO3D objects | diversity: 5.04 -> 5.91 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- ContactOpt — ContactOpt: Optimizing Contact To Improve Grasps (CVPR 2021) · better on 3 of 3 shared results
- ContactGen — ContactGen: Generative Contact Modeling for Grasp Generation (ICCV 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Text2HOI: Text-guided 3D Motion Generation for Hand-Object Interaction (CVPR 2024 · similarity 0.90)
- VolCo: Volumetric Contact for High-Fidelity Human Grasp Generation (NeurIPS 2026 · similarity 0.89)
- DiffGrasp: Whole-Body Grasping Synthesis Guided by Object Motion Using a Diffusion Model (AAAI 2025 · similarity 0.88)
- Language-driven Grasp Detection (CVPR 2024 · similarity 0.88)
- How Do I Do That? Synthesizing 3D Hand Motion and Contacts for Everyday Interactions (CVPR 2025 · similarity 0.88)
- G-HOP: Generative Hand-Object Prior for Interaction Reconstruction and Grasp Synthesis (CVPR 2024 · similarity 0.88)
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.
