Research map / Video / Video segmentation and tracking
Learning Shadow Correspondence for Video Shadow Detection
Xinpeng Ding, Jingwen Yang, Xiaowei Hu, Xiaomeng Li
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 SC-Cor to DSD (ablation) | DSD · ViSha | BER: 19.89 -> 15.67 | Improved |
| add SC-Cor to DSD (ablation) | DSD · ViSha | TS: 74.68 -> 77.09 | Improved |
| add brightness shift to DSD (ablation) | DSD · ViSha | TS: 74.68 -> 74.92 | No clear effect |
| add brightness shift to DSD (ablation) | DSD · ViSha | BER: 19.89 -> 16.82 | Improved |
| add brightness shift to SC-Cor (ablation) | SC-Cor · ViSha | AVG: 67.00 -> 68.22 | Improved |
| add bidirectional correspondence (ablation) | SC-Cor · ViSha | AVG: 67.17 -> 68.22 | Improved |
| add non-shadow regularization to shadow-only (ablation) | SC-Cor · ViSha | AVG: 67.04 -> 68.22 | Improved |
| add shadow regularization to non-shadow-only (ablation) | SC-Cor · ViSha | AVG: 66.91 -> 68.22 | Improved |
| use 4 frames instead of 2 (ablation) | SC-Cor · ViSha | AVG: 68.01 -> 68.39 | Improved |
| increase frame interval from 1 to 7 (ablation) | SC-Cor · ViSha | TS: 76.94 -> 78.23 | Mixed |
| increase frame interval from 1 to 7 (ablation) | SC-Cor · ViSha | BER: 14.51 -> 15.64 | Mixed |
| plug SC-Cor into DSD | DSD · ViSha | AVG: 63.29 -> 68.22 | Improved |
Closest papers by meaning
- Triple-Cooperative Video Shadow Detection (CVPR 2021 · similarity 0.90)
- Video Shadow Detection via Spatio-Temporal Interpolation Consistency Training (CVPR 2022 · similarity 0.89)
- DTTNet: Improving Video Shadow Detection via Dark-Aware Guidance and Tokenized Temporal Modeling (AAAI 2026 · similarity 0.89)
- Two-Stage Video Shadow Detection via Temporal-Spatial Adaption (ECCV 2024 · similarity 0.88)
- SCOTCH and SODA: A Transformer Video Shadow Detection Framework (CVPR 2023 · similarity 0.88)
- MetaShadow: Object-Centered Shadow Detection, Removal, and Synthesis (CVPR 2025 · similarity 0.84)
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.
