Research map / Detection and segmentation / Weakly supervised detection and segmentation
Learning Self-supervised Low-Rank Network for Single-Stage Weakly and Semi-supervised Semantic Segmentation
Junwen Pan, Pengfei Zhu, Kaihua Zhang, Bing Cao, Yu Wang, Dingwen Zhang, Junwei Han, Qinghua 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 |
|---|---|---|---|
| use two views instead of a single view (ablation) | SLRNet · Pascal VOC train | mIoU (%): 60.88 -> 64.07 | Improved |
| use different scales (0.5,1.0) instead of the same scale (1.0,1.0) (ablation) | SLRNet · Pascal VOC train | mIoU (%): 61.53 -> 64.07 | Improved |
| add a third view at scale 1.5 (ablation) | SLRNet · Pascal VOC train | mIoU (%): 64.07 -> 63.90 | No clear effect |
| use independent color/flip transformations across views (ablation) | SLRNet · Pascal VOC train | mIoU (%): 63.99 -> 64.07 | No clear effect |
| add the CVLR module (ablation) | SLRNet · Pascal VOC train | mIoU (%): 61.11 -> 63.99 | Improved |
| share one dictionary across views instead of separate dictionaries (ablation) | SLRNet · Pascal VOC train | mIoU (%): 61.24 -> 63.99 | Improved |
| add latent space regularization (ablation) | SLRNet · Pascal VOC train | mIoU (%): 62.83 -> 63.99 | Improved |
| use CVLR instead of EMAU (ablation) | SLRNet · Pascal VOC train | mIoU (%): 61.76 -> 63.99 | Improved |
| use CVLR instead of Hamburger (ablation) | SLRNet · Pascal VOC train | mIoU (%): 59.75 -> 63.99 | Improved |
| increase CVLR iterations beyond one step (ablation) | SLRNet · Pascal VOC train | mIoU (%) | No clear effect |
| add CRF post-processing to pseudo-masks | SLRNet · Pascal VOC train | mIoU (%): 67.1 -> 70.3 | Improved |
| distill into DeeplabV3+ with SLRNet pseudo-labels | SLRNet · Pascal VOC val | mIoU (%): 67.2 -> 69.3 | Improved |
Closest papers by meaning
- Single-Stage Semantic Segmentation From Image Labels (CVPR 2020 · similarity 0.91)
- Self Correspondence Distillation for End-to-End Weakly-Supervised Semantic Segmentation (AAAI 2023 · similarity 0.91)
- Progressive Feature Self-Reinforcement for Weakly Supervised Semantic Segmentation (AAAI 2024 · similarity 0.90)
- Label-Efficient Few-Shot Semantic Segmentation with Unsupervised Meta-Training (AAAI 2024 · similarity 0.90)
- From SAM to CAMs: Exploring Segment Anything Model for Weakly Supervised Semantic Segmentation (CVPR 2024 · similarity 0.89)
- W2P: Switching from Weak Supervision to Partial Supervision for Semantic Segmentation (AAAI 2024 · similarity 0.89)
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.
