Research map / Label-efficient and robust learning / Recognition and long-tail
Disentangled Hypergraph-Guided Mamba Scanning for Fine-Grained Visual Recognition
Zhongwei Xiong, Hao Wang, Xiaoyan Yu, Lingling Li, Xuezhuan Zhao, Taisong Jin
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 hypergraph and disentangled scanning modules to VMamba | VMamba · CUB-200-2011 | top-1 accuracy (%): 89.7 -> 92.9 | Improved |
| add hypergraph convolution (HSA) (ablation) | DHMamba · CUB-200-2011 | top-1 accuracy (%): 91.7 -> 92.9 | Improved |
| add local information gradient scanning (ablation) | DHMamba · CUB-200-2011 | top-1 accuracy (%): 91.5 -> 92.9 | Improved |
| add disentangled scanning (ablation) | DHMamba · CUB-200-2011 | top-1 accuracy (%): 91.0 -> 92.9 | Improved |
| raise number of clusters k from 5 to 20 (ablation) | DHMamba HSA · CUB-200-2011 | top-1 accuracy (%): 92.68 -> 92.92 | Improved |
| raise number of clusters k from 20 to 25 (ablation) | DHMamba HSA · CUB-200-2011 | top-1 accuracy (%): 92.92 -> 92.71 | Hurt |
| use DHMamba over VMamba | VMamba · NABirds | top-1 accuracy (%): 88.7 -> 92.0 | Improved |
| use DHMamba over VMamba | VMamba · Stanford Cars | top-1 accuracy (%): 94.5 -> 96.7 | Improved |
| use DHMamba over HI2R | HI2R · CUB-200-2011 | top-1 accuracy (%): 92.5 -> 92.9 | Improved |
| scale input resolution from 320x320 to 448x448 | DHMamba vs ViT · CUB-200-2011 | training time multiplier: 1.77 -> 1.50 | Improved |
Closest papers by meaning
- Vision Mamba Mender (NeurIPS 2024 · similarity 0.86)
- QuadMamba: Learning Quadtree-based Selective Scan for Visual State Space Model (NeurIPS 2024 · similarity 0.86)
- Multi-Scale VMamba: Hierarchy in Hierarchy Visual State Space Model (NeurIPS 2024 · similarity 0.86)
- MambaOut: Do We Really Need Mamba for Vision? (CVPR 2025 · similarity 0.85)
- Mamba-Adaptor: State Space Model Adaptor for Visual Recognition (CVPR 2025 · similarity 0.85)
- Demystify Mamba in Vision: A Linear Attention Perspective (NeurIPS 2024 · 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.
