Research map / Video / Video segmentation and tracking
Every Frame Counts: Joint Learning of Video Segmentation and Optical Flow
Read the paperCheck 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 EFC joint training to PSPNet | PSPNet · Cityscapes test | class IoU: 78.4 -> 80.2 | Helped |
| Add EFC joint training to Dilation8 | Dilation8 · CamVid test | mIoU: 65.3 -> 67.4 | Helped |
| Jointly train segmentation and optical flow with EFC | FlowNetS flow branch · KITTI 2015 flow training set | EPE (Noc): 4.30 -> 3.93 | Helped |
| Add temporal consistency using fixed pre-trained FlowNetS (ablation) | ResNet50 + PSPNet · Cityscapes validation | IoU: 76.20 -> 77.02 | Helped |
| Jointly learn flow rather than use fixed pre-trained FlowNetS (ablation) | TCC single-pair segmentation · Cityscapes validation | IoU: 77.02 -> 77.58 | Helped |
| Add an occlusion mask to single-pair TCC (ablation) | TCC_single · Cityscapes validation | IoU: 77.58 -> 77.79 | Helped |
| Use five randomly selected frame pairs instead of a single pair (ablation) | TCC_single + OM · Cityscapes validation | IoU: 77.79 -> 78.07 | Helped |
| Add unlabeled data (ablation) | TCC_multi + OM · Cityscapes validation | IoU: 78.07 -> 78.44 | Helped |
| Add occlusion estimation (ablation) | UL · KITTI | EPE (Noc): 7.53 -> 7.23 | Helped |
| Add segmentation branch and temporal consistency (ablation) | UL · KITTI | EPE (Noc): 7.53 -> 4.94 | Helped |
| Add occlusion estimation to UL + TC (ablation) | UL + TC · KITTI | EPE (Noc): 4.94 -> 4.51 | Helped |
| Add full occlusion handling (ablation) | UL + OE + TC · KITTI | EPE (Noc): 4.51 -> 3.93 | Helped |
Closest papers by meaning
- Video Segmentation via Object Flow (CVPR 2016 · similarity 0.91)
- Optical Flow With Semantic Segmentation and Localized Layers (CVPR 2016 · similarity 0.90)
- Semantic Flow: Learning Semantic Fields of Dynamic Scenes from Monocular Videos (ICLR 2024 · similarity 0.89)
- Learning Spatial-Semantic Features for Robust Video Object Segmentation (ICLR 2025 · similarity 0.88)
- Exploring Temporal Feature Correlation for Efficient and Stable Video Semantic Segmentation (AAAI 2024 · similarity 0.87)
- VidSeg: Training-free Video Semantic Segmentation based on Diffusion Models (CVPR 2025 · similarity 0.87)
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.