Research map / 3D vision / 3D generation
Geo-DPO: Aligning Semantic Intent with Geometry for 3D Affordance Segmentation
Ziqian Yang, Xiaolei Wang, Xianglin Qiu, Weiguang Zhao, Quan Zhang, Jimin Xiao
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 HGA (ablation) | baseline ShapeLLM-based model · SeqAfford Sequential | mIoU: 14.5 -> 14.7 | Improved |
| add CAPO (ablation) | baseline ShapeLLM-based model · SeqAfford Sequential | mIoU: 14.5 -> 14.8 | Improved |
| add both HGA and CAPO (ablation) | baseline ShapeLLM-based model · SeqAfford Sequential | mIoU: 14.5 -> 15.1 | Improved |
| add CAPO on top of HGA (ablation) | HGA model · SeqAfford Sequential | mIoU: 14.7 -> 15.1 | Improved |
| raise CAPO weight lambda from 0.0 to 0.05 (ablation) | Geo-DPO · SeqAfford Sequential | mIoU: 14.7 -> 14.7 | No clear effect |
| raise CAPO weight lambda from 0.2 to 0.4 (ablation) | Geo-DPO · SeqAfford Sequential | mIoU: 15.1 -> 14.8 | Hurt |
| use cross-attention instead of addition for HGA fusion (ablation) | HGA · SeqAfford Sequential | mIoU: 14.4 -> 15.1 | Improved |
| use cross-attention instead of concatenation for HGA fusion (ablation) | HGA · SeqAfford Sequential | mIoU: 14.5 -> 15.1 | Improved |
| replace SeqAfford with Geo-DPO | SeqAfford · SeqAfford Seen | mIoU: 19.5 -> 20.9 | Improved |
| replace SeqAfford with Geo-DPO | SeqAfford · SeqAfford Unseen | mIoU: 13.8 -> 14.8 | Improved |
| replace SeqAfford with Geo-DPO | SeqAfford · SeqAfford Sequential | mIoU: 14.6 -> 15.1 | Improved |
| replace SeqAfford with Geo-DPO | SeqAfford · SeqAfford Unseen | AUC: 82.4 -> 85.6 | Improved |
Closest papers by meaning
- SeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language Model (CVPR 2025 · similarity 0.91)
- 3D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D Worlds (ICLR 2025 · similarity 0.90)
- LASO: Language-guided Affordance Segmentation on 3D Object (CVPR 2024 · similarity 0.89)
- UniFunc3D: Unified Active Spatial-Temporal Grounding for 3D Affordance Segmentation (NeurIPS 2026 · similarity 0.89)
- GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance Grounding (CVPR 2025 · similarity 0.88)
- GEAL: Generalizable 3D Affordance Learning with Cross-Modal Consistency (CVPR 2025 · similarity 0.87)
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.
