Research map / Detection and segmentation / Open-set and few-shot detection
Intermediate Connectors and Geometric Priors for Language-Guided Affordance Segmentation on Unseen Object Categories
Yicong Li, Yiyang Chen, Zhenyuan Ma, Junbin Xiao, Xiang Wang, Angela Yao
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 cross-modal connector (CMC) and GAQG to baseline (ablation) | GLANCE · LASO (Seen) | mIoU: 21.3 -> 23.3 | Improved |
| add CMC (ablation) | GLANCE (with GAQG) · LASO (Seen) | mIoU: 22.3 -> 23.3 | Improved |
| add GAQG (ablation) | GLANCE (with CMC) · LASO (Seen) | mIoU: 21.7 -> 23.3 | Improved |
| add intra-view reliability sampling (ablation) | GLANCE GAQG · LASO (Seen) | mIoU: 22.4 -> 23.3 | Improved |
| add cross-view consistency (ablation) | GLANCE GAQG · LASO (Seen) | mIoU: 21.9 -> 23.3 | Improved |
| use GAQG 2D-prior key points instead of random sampler (ablation) | GLANCE · LASO (Seen) | mIoU: 21.5 -> 23.3 | Improved |
| use GAQG 2D-prior queries instead of point cloud feature sampler (ablation) | GLANCE · LASO (Seen) | mIoU: 21.9 -> 23.3 | Improved |
| embed CMC in last 4 blocks (9-12) instead of last 2 (ablation) | GLANCE · LASO (Seen) | mIoU: 22.2 -> 23.3 | Improved |
| embed CMC in last 4 blocks (9-12) instead of last 8 (ablation) | GLANCE · LASO (Seen) | mIoU: 22.1 -> 23.3 | Improved |
| use CMC hidden dimension 64 instead of 128 (ablation) | GLANCE · LASO (Seen) | mIoU: 21.7 -> 23.3 | Improved |
| use 4 views instead of 1 view (ablation) | GLANCE · LASO (Unseen) | mIoU: 18.0 -> 18.7 | Improved |
| use 4 views instead of 2 views (ablation) | GLANCE · LASO (Seen) | mIoU: 21.9 -> 23.3 | Improved |
Closest papers by meaning
- LASO: Language-guided Affordance Segmentation on 3D Object (CVPR 2024 · similarity 0.93)
- 3D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D Worlds (ICLR 2025 · similarity 0.92)
- SeqAfford: Sequential 3D Affordance Reasoning via Multimodal Large Language Model (CVPR 2025 · similarity 0.91)
- Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic Priors (ICLR 2025 · similarity 0.90)
- One-Shot Open Affordance Learning with Foundation Models (CVPR 2024 · similarity 0.90)
- UniFunc3D: Unified Active Spatial-Temporal Grounding for 3D Affordance Segmentation (NeurIPS 2026 · similarity 0.89)
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.
