Research map / Detection and segmentation / Open-set and few-shot detection
Multispectral Video Semantic Segmentation: A Benchmark Dataset and Baseline
Wei Ji, Jingjing Li, Cheng Bian, Zongwei Zhou, Jiaying Zhao, Alan L. Yuille, Li Cheng
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 multispectral video contexts using MVNet | DeepLabv3+ · MVSeg test | mIoU: 51.59%→54.52% | Improved |
| add the TIR stream with direct fusion (ablation) | DeepLabv3+ on RGB input · MVSeg test | mIoU: 51.59%→52.53% | Improved |
| replace direct fusion with a cascaded decoder (ablation) | RGB+TIR segmentation · MVSeg test | mIoU: 52.53%→52.87% | Improved |
| use prototypical memory and all-to-prototype attention instead of dense or neighborhood pixel memory (ablation) | MVNet · MVSeg test | mIoU: 52.87%→54.03% | Improved |
| add the full MVRegulator, including cross-spectral contrast (ablation) | MVNet with MVFuseproto · MVSeg test | mIoU: 54.03%→54.52% | Improved |
| remove cross-spectral contrast from MVRegulator (ablation) | MVNet with MVFuseproto · MVSeg test | mIoU: 54.26%→54.52% | Hurt |
| increase memory frames from 0 to 3 (ablation) | MVNet · MVSeg test | mIoU: 52.87%→54.52% | Improved |
| increase memory frames beyond 3 (ablation) | MVNet · MVSeg test | mIoU: 54.52%→54.57%→54.52% | No clear effect |
| use memory sample rate S = 3 rather than other tested rates (ablation) | MVNet with M = 3 · MVSeg test | mIoU: 54.25; 54.47; 54.52; 54.41; 54.39 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MFNet — MFNet: Multi-Filter Directive Network for Weakly Supervised Salient Object Detection (ICCV 2021) · better on 2 of 2 shared results
- CCNet — CCNet: Criss-Cross Attention for Semantic Segmentation (ICCV 2019) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Unleashing Multispectral Video's Potential in Semantic Segmentation: A Semi-supervised Viewpoint and New UAV-View Benchmark (NeurIPS 2024 · similarity 0.93)
- UNIP: Rethinking Pre-trained Attention Patterns for Infrared Semantic Segmentation (ICLR 2025 · similarity 0.87)
- VidSeg: Training-free Video Semantic Segmentation based on Diffusion Models (CVPR 2025 · similarity 0.86)
- Weakly Supervised Semantic Segmentation for Driving Scenes (AAAI 2024 · similarity 0.86)
- Hierarchy-Agnostic Unsupervised Segmentation: Parsing Semantic Image Structure (NeurIPS 2024 · similarity 0.86)
- Integrating Low-Level Visual Cues for Enhanced Unsupervised Semantic Segmentation (AAAI 2025 · similarity 0.86)
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.
