AI Paper Atlas

Research map / Detection and segmentation / CNN features

SF-Mamba: Rethinking State Space Model for Vision

Masakazu Yoshimura, Teruaki Hayashi, Yuki Hoshino, Wei-Yao Wang, Takeshi Ohashi

ICML 2026 · Detection and segmentation › CNN features · approach: Vision state space models · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
add auxiliary token swapping (ablation)MambaVision-T · ImageNet-1K / ADE20Ktop-1 / mIoU: 82.2 -> 82.5 (IN1K); 46.0 -> 47.2 (ADE20K)Improved
add learnable auxiliary tokens without swapping (ablation)MambaVision-T · ImageNet-1K / ADE20Ktop-1 / mIoU: 82.2 -> 82.1 (IN1K); 46.0 -> 46.2 (ADE20K)Mixed
initialize aux tokens as global average instead of learnable (ablation)SF-Mamba-T · ImageNet-1K / ADE20Ktop-1 / mIoU: 82.3 -> 82.4 (IN1K); 46.5 -> 46.8 (ADE20K)Improved
discard aux token after first attention layer instead of before attention (ablation)SF-Mamba-T · ADE20KmIoU: 46.8 -> 47.2Improved
discard aux token after all attention instead of after first attention (ablation)SF-Mamba-T · ADE20KmIoU: 47.2 -> 46.6Hurt
add batch folding (BFold) with B1=1 (ablation)MambaVision-T · ImageNet-1K throughputimg/s: 6989 -> 7601Improved
use adaptive B1 with folding-compatible 1-D conv (ablation)MambaVision-T · ImageNet-1K throughputimg/s: 7279 -> 7685Improved
adaptive B1 without compatible conv vs fixed B1=4 (ablation)MambaVision-T · ImageNet-1K throughputimg/s: 7641 -> 7279Hurt
batch folding of SSM kernel at short sequences (ablation)Mamba SSM kernel · SSM kernel speedspeedup: 110% to 180%Improved
replace uni-scan with parallel bi-scan (add) (ablation)MambaVision-T · ImageNet-1Ktop-1 / img/s: 82.2 -> 82.3; 6979 -> 6235 img/sMixed
replace uni-scan with series bi-scan (ablation)MambaVision-T · ImageNet-1Ktop-1: 82.2 -> 82.3No clear effect
replace uni-scan with swap on Mamba-only macro-architecture (ablation)MambaVision-T w/o Attention · ImageNet-1Ktop-1: 80.2 -> 81.0Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy