Research map / Video / Video-language understanding
Video Mamba Suite: State Space Model as a Versatile Alternative for Video Understanding
Guo Chen, Yifei Huang, Jilan Xu, Baoqi Pei, Jiahao Wang, Zhe Chen, Zhiqi Li, Tong Lu, Limin Wang
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 |
|---|---|---|---|
| replace window attention with ViM block | ActionFormer · HACS Segment | mAP@Avg: 43.34 -> 44.26 | Improved |
| use DBM block instead of ViM block | ActionMamba · HACS Segment | mAP@Avg: 44.26 -> 44.56 | Improved |
| use DBM block instead of ViM block | ActionMamba · FineAction | mAP@Avg: 28.36 -> 29.04 | Improved |
| use DBM instead of ViM | ASMamba · Breakfast | Accuracy: 75.8 -> 73.5 | Hurt |
| use DBM instead of ViM | ASMamba · 50Salads | Accuracy: 85.4 -> 83.7 | Hurt |
| use encoder-only Mamba instead of encoder-decoder ASFormer | ASMamba · 50Salads | Accuracy: 85.6 -> 85.4 | Hurt |
| use encoder-only ASFormer instead of encoder-decoder | ASFormer · GTEA | Accuracy: 77.1 -> 75.4 | Hurt |
| replace deformable attention encoder with DBM | PDVC · YouCook2 dense captioning | CIDEr: 20.48 -> 22.11 | Improved |
| replace ViM with DBM | PDVC · ActivityNet dense captioning | SODA: 5.33 -> 5.27 | Hurt |
| replace short attention with causal Mamba | Testra · EK-100 anticipation | Action recall@5: 14.1 -> 15.2 | Improved |
| replace Transformer with Mamba block as cross-modal encoder | UniVTG · QVHighlights | mAP Avg: 38.48 -> 44.74 | Improved |
| replace Transformer with Mamba block | UniVTG · Charade-STA | R1@0.5: 57.26 -> 57.07 | No clear effect |
Closest papers by meaning
- MambaOut: Do We Really Need Mamba for Vision? (CVPR 2025 · similarity 0.92)
- Vision Mamba Mender (NeurIPS 2024 · similarity 0.91)
- Mamba-Adaptor: State Space Model Adaptor for Visual Recognition (CVPR 2025 · similarity 0.90)
- MoMa: Modulating Mamba for Adapting Image Foundation Models to Video Recognition (ICML 2025 · similarity 0.89)
- Demystify Mamba in Vision: A Linear Attention Perspective (NeurIPS 2024 · similarity 0.89)
- Achilles' Heel of Mamba: Essential difficulties of the Mamba architecture demonstrated by synthetic data (NeurIPS 2025 · similarity 0.89)
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.
