Research map / Detection and segmentation / Weakly supervised detection and segmentation
Competing for Pixels: A Self-Play Algorithm for Weakly-Supervised Semantic Segmentation
Shaheer Ullah Saeed, Shiqi Huang, João Ramalhinho, Iani J. M. B. Gayo, Nina Montaña-Brown, Ester Bonmati, Stephen P. Pereira, Brian Ritchie Davidson, Dean C. Barratt, Matthew J. Clarkson, Yipeng Hu
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 self-play opponent agent instead of fixed 0.8 threshold (ablation) | RLSP · VOC val | mIoU: 70.3 -> 78.9 | Improved |
| add repetition and iteration shaping rewards (ablation) | RLSP · VOC val | mIoU: 75.7 -> 78.9 | Improved |
| add iteration-bounding reward on top of repetition reward (ablation) | RLSP · VOC val | mIoU: 76.4 -> 78.9 | Improved |
| add repetition reward on top of iteration reward (ablation) | RLSP · VOC val | mIoU: 77.7 -> 78.9 | Improved |
| add shaping rewards without self-play (ablation) | non-self-play variant · VOC val | mIoU: 68.8 -> 70.3 | Improved |
| increase patch size from 2x2 to 6x6 (ablation) | RLSP · VOC val | mIoU: 76.4 -> 78.9 | Improved |
| increase patch size beyond 6x6 to 8x8 (ablation) | RLSP · VOC val | mIoU: 78.9 -> 78.3 | No clear effect |
| use self-play RL instead of sliding-window with 0.5 threshold | object presence detector · VOC val | mIoU: 64.9 -> 78.9 | Improved |
| lower non-self-play selection threshold from 0.8 to 0.2 | non-SP · VOC val | mIoU: 70.3 -> 68.5 | Hurt |
| switch backbone from ResNet38 to ResNet101 or VGG16 (ablation) | RLSP · VOC val | mIoU: reduced performance by 0.5 or 1.2 mIoU points | No clear effect |
| RLSP versus best prior MARS | RLSP · CMPS (RE) | mDSC: 54.2 -> 58.8 | Improved |
| RLSP versus best prior MARS | RLSP · VOC test | mIoU: 77.2 -> 78.7 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- USAGE — USAGE: A Unified Seed Area Generation Paradigm for Weakly Supervised Semantic Segmentation (ICCV 2023) · better on 3 of 3 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
- Progressive Feature Self-Reinforcement for Weakly Supervised Semantic Segmentation (AAAI 2024 · similarity 0.91)
- Threshold Matters in WSSS: Manipulating the Activation for the Robust and Accurate Segmentation Model Against Thresholds (CVPR 2022 · similarity 0.91)
- Universal Weakly Supervised Segmentation by Pixel-to-Segment Contrastive Learning (ICLR 2021 · similarity 0.90)
- Separate and Conquer: Decoupling Co-occurrence via Decomposition and Representation for Weakly Supervised Semantic Segmentation (CVPR 2024 · similarity 0.89)
- W2P: Switching from Weak Supervision to Partial Supervision for Semantic Segmentation (AAAI 2024 · similarity 0.88)
- Weakly Supervised Semantic Segmentation for Driving Scenes (AAAI 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.
