Research map / Video / Video segmentation and tracking
Video Shadow Detection via Spatio-Temporal Interpolation Consistency Training
Check 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 |
|---|---|---|---|
| add Feature Fusion Module to the encoder-decoder network (ablation) | SANet · DS | IoU: 0.433 -> 0.492 | Helped |
| add the refiner to ED+FFM (ablation) | SANet · DS | Fβ: 0.616 -> 0.660 | Helped |
| add the Detail Attentive Module to ED+FFM+R (ablation) | SANet · DS | Fβ: 0.660 -> 0.708 | Helped |
| add scale consistency to the basic SANet (ablation) | STICT · DS | Fβ: 0.501 -> 0.518 | Helped |
| add temporal interpolation consistency to B+SC (ablation) | STICT · DS | MAE: 0.092 -> 0.079 | Helped |
| add spatial interpolation consistency to B+SC+TIC (ablation) | STICT · DS | IoU: 0.320 -> 0.370 | Helped |
| use random interpolation in RGB space instead of feature-space spatial interpolation (ablation) | Spatial ICT · DS | Fβ: 0.646 -> 0.447 | Hurt |
| replace feature-space spatial interpolation with feature-space random interpolation (ablation) | Spatial ICT · DS | Fβ: 0.646 -> 0.557 | Hurt |
| transfer image-trained SANet to video without consistency training (ablation) | STICT · DS | MAE: 0.028 -> 0.093 | Hurt |
| train with unlabeled video frames using STICT rather than direct image-to-video transfer | SANet · DS | Fβ: 0.501 -> 0.646 | Helped |
Closest papers by meaning
- Triple-Cooperative Video Shadow Detection (CVPR 2021 · similarity 0.88)
- A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection (CVPR 2020 · similarity 0.86)
- Supervision Interpolation via LossMix: Generalizing Mixup for Object Detection and Beyond (AAAI 2024 · similarity 0.83)
- Stable Mean Teacher for Semi-supervised Video Action Detection (AAAI 2025 · similarity 0.83)
- Generalizing Deepfake Video Detection with Plug-and-Play: Video-Level Blending and Spatiotemporal Adapter Tuning (CVPR 2025 · similarity 0.82)
- ZOOM: Learning Video Mirror Detection with Extremely-Weak Supervision (AAAI 2024 · similarity 0.82)
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.