Research map / Video / Video segmentation and tracking
Deep Spatio-Temporal Random Fields for Efficient Video Segmentation
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 3-frame VideoGCRF | ResNet-101 segmentation model · CamVid test, no Cityscapes pretraining | m-IoU: 65.7 -> 67.0 | Helped |
| Add 3-frame VideoGCRF | Cityscapes-pretrained ResNet-101 segmentation model · CamVid test | m-IoU: 73.3 -> 75.2 | Helped |
| Add temporal VideoGCRF with longer frame context | ResNet-50 Mask R-CNN · DAVIS Person val | AP: 0.321 -> 0.331 -> 0.330 -> 0.349 | Helped |
| Add 3-frame VideoGCRF | Online Adaptation · DAVIS val | mean IoU: 85.6 -> 86.5 | Helped |
| Increase spatial and temporal embedding dimensions from the base-net setting to 128 (ablation) | VideoGCRF · DAVIS Person, 2-frame fully connected | mean IoU: 81.16 -> 86.38 | Helped |
| Vary spatial and temporal embedding dimensions (ablation) | VideoGCRF · DAVIS Person, 2-frame fully connected | mean IoU: 84.89; 85.21; 85.20; 84.98; 85.18; 86.38; 86.34; 84.91; 85.92; 86.37; 85.95; 84.92; 84.85; 85.95; 84.95; 84.21 | Mixed |
| Increase temporal context from 2 to 7 frames (ablation) | VideoGCRF instance segmentation · DAVIS Person val | AP: 0.331 -> 0.349 | Helped |
| Vary temporal neighbourhood and context (ablation) | VideoGCRF semantic segmentation · DAVIS Person | mean IoU: 86.38; 86.42; 86.51; 86.70; 86.82; 86.98; 86.79; 86.82; 86.42 | Mixed |
Closest papers by meaning
- Gaussian Conditional Random Field Network for Semantic Segmentation (CVPR 2016 · similarity 0.86)
- CNN in MRF: Video Object Segmentation via Inference in a CNN-Based Higher-Order Spatio-Temporal MRF (CVPR 2018 · similarity 0.86)
- Feature Space Optimization for Semantic Video Segmentation (CVPR 2016 · similarity 0.86)
- Exploiting Temporal State Space Sharing for Video Semantic Segmentation (CVPR 2025 · similarity 0.85)
- LASER: A Neuro-Symbolic Framework for Learning Spatio-Temporal Scene Graphs with Weak Supervision (ICLR 2025 · similarity 0.85)
- LiVOS: Light Video Object Segmentation with Gated Linear Matching (CVPR 2025 · similarity 0.85)
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.