Research map / Detection and segmentation / Weakly supervised detection and segmentation
LOCATE: Localize and Transfer Object Parts for Weakly Supervised Affordance Grounding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use regional rather than global knowledge transfer (ablation) | LOCATE · AGD20K seen and unseen | KLD: 1.732 -> 1.516 (seen); 1.971 -> 1.823 (unseen) | Helped |
| Add concentration loss to regional transfer (ablation) | LOCATE · AGD20K seen and unseen | KLD: 1.516 -> 1.491 (seen); 1.823 -> 1.750 (unseen) | Helped |
| Add S and cosine embedding loss for part-level transfer (ablation) | LOCATE with regional transfer · AGD20K seen and unseen | KLD: 1.516 -> 1.236 (seen); 1.823 -> 1.439 (unseen) | Helped |
| Add concentration loss after adding S and cosine embedding loss (ablation) | LOCATE · AGD20K seen and unseen | KLD: 1.236 -> 1.226 (seen); 1.439 -> 1.405 (unseen) | Helped |
| Use DINO-ViT rather than DINO-ResNet-50 features with S (ablation) | LOCATE · AGD20K unseen | KLD: 1.707 -> 1.405 | Helped |
| Add S to DINO-ResNet-50 features (ablation) | LOCATE · AGD20K unseen | KLD: 1.701 -> 1.707 | Hurt |
| Add S to DINO-ResNet-50 features (ablation) | LOCATE · AGD20K seen | KLD: 1.482 -> 1.449 | Helped |
| Increase the number of exocentric images from 1 to 3 (ablation) | LOCATE · AGD20K seen and unseen | affordance grounding performance | Helped |
| Use five rather than three prototypes (ablation) | LOCATE · AGD20K unseen | affordance grounding performance | Helped |
| Use three rather than five prototypes (ablation) | LOCATE · AGD20K seen | affordance grounding performance | Helped |
| Use LOCATE rather than adapted AffCorrs | AGD20K affordance grounding · AGD20K seen and unseen | KLD: 1.407 -> 1.226 (seen); 1.618 -> 1.405 (unseen) | Helped |
Closest papers by meaning
- Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic Priors (ICLR 2025 · similarity 0.93)
- Weakly Supervised Multimodal Affordance Grounding for Egocentric Images (AAAI 2024 · similarity 0.92)
- Reasoning Mamba: Hypergraph-Guided Region Relation Calculating for Weakly Supervised Affordance Grounding (CVPR 2025 · similarity 0.91)
- Grounding 3D Object Affordance with Language Instructions, Visual Observations and Interactions (CVPR 2025 · similarity 0.90)
- Learning 2D Invariant Affordance Knowledge for 3D Affordance Grounding (AAAI 2025 · similarity 0.90)
- GREAT: Geometry-Intention Collaborative Inference for Open-Vocabulary 3D Object Affordance Grounding (CVPR 2025 · 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.