Research map / Detection and segmentation / Weakly supervised detection and segmentation
COHESIV: Contrastive Object and Hand Embedding Segmentation In Video
Dandan Shan, Richard Ely Locke Higgins, David Fouhey
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 embedding contrastive loss to attention-only training (ablation) | COHESIV · 100DOH All | mIoU (%): 42.8 -> 51.9 | Improved |
| add embedding contrastive loss to attention-only training (ablation) | COHESIV · EPIC-KITCHENS All | mIoU (%): 38.1 -> 43.2 | Improved |
| add responsibility attention loss to embeddings-only training (ablation) | COHESIV · 100DOH All | mIoU (%): 25.7 -> 51.9 | Improved |
| use SE-ResNeXt backbone instead of ResNet (ablation) | COHESIV · 100DOH All | mIoU (%): 45.8 -> 51.9 | Improved |
| use SE-ResNeXt backbone instead of ResNet (ablation) | COHESIV · EPIC-KITCHENS Hand | mIoU (%): 55.2 -> 60.7 | Improved |
| use external hand detector query point instead of predicted query (ablation) | COHESIV · 100DOH All | mIoU (%): 47.7 -> 51.9 | Improved |
| use external hand detector query point instead of predicted query (ablation) | COHESIV · EPIC-KITCHENS Obj | mIoU (%): 19.4 -> 19.5 | No clear effect |
| learn network from responsibility pseudo-labels instead of thresholding them directly | COHESIV · 100DOH Pair | mIoU (%): 37.0 -> 46.5 | Improved |
| learn network from responsibility pseudo-labels instead of thresholding them directly | COHESIV · EPIC-KITCHENS Pair | mIoU (%): 30.0 -> 42.1 | Improved |
| weakly supervised COHESIV vs fully supervised bounding box detector | COHESIV · 100DOH All | mIoU (%): 56.9 -> 51.9 | Hurt |
| weakly supervised COHESIV vs fully supervised bounding box detector | COHESIV · EPIC-KITCHENS Hand | mIoU (%): 53.8 -> 60.7 | Improved |
Later papers that compared with this method
1 later accepted paper reports COHESIV as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- MOVES: Manipulated Objects in Video Enable Segmentation (CVPR 2023 · better on 1 of 1 shared result)
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
- MOVES: Manipulated Objects in Video Enable Segmentation (CVPR 2023 · similarity 0.89)
- ComPose: When to Trust Hands for Object Pose Tracking (NeurIPS 2026 · similarity 0.85)
- SiMHand: Mining Similar Hands for Large-Scale 3D Hand Pose Pre-training (ICLR 2025 · similarity 0.85)
- Modeling Fine-Grained Hand-Object Dynamics for Egocentric Video Representation Learning (ICLR 2025 · similarity 0.85)
- HOIST-Former: Hand-held Objects Identification Segmentation and Tracking in the Wild (CVPR 2024 · similarity 0.85)
- HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video (CVPR 2024 · similarity 0.85)
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.
