Research map / Detection and segmentation / CNN features
AKCMamba-YOLO: Selective State Space Models For Real-Time Object Detection
Long Chen, Hui Wang, Man Xu, Zexuan Li, Zizhu Fan
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 C2f with 3CAKC in the backbone (ablation) | YOLOv8-N · Power Tower Foreign Object | AP50:95: 67.4 -> 70.5 | Improved |
| replace C2f with 3CAKC in the backbone (ablation) | YOLOv8-N · Power Tower Foreign Object | precision: 89.7 -> 87.2 | Hurt |
| add AKSS2D state space module to the backbone (ablation) | 3CAKC backbone · Power Tower Foreign Object | AP50:95: 70.5 -> 73.0 | Improved |
| add AKCAttention to the backbone (ablation) | 3CAKC+AKSS2D backbone · Power Tower Foreign Object | AP50:95: 73.0 -> 75.0 | Improved |
| replace neck C2f with 4CAKC (ablation) | YOLOv8-N · Power Tower Foreign Object | AP50:95: 67.4 -> 70.6 | Improved |
| replace neck C2f with 4CAKC (ablation) | YOLOv8-N · Power Tower Foreign Object | precision: 89.7 -> 87.4 | Hurt |
| add AKSS2D to the neck (ablation) | 4CAKC neck · Power Tower Foreign Object | AP50:95: 70.6 -> 71.9 | Improved |
| add AKCAttention to the neck (ablation) | 4CAKC+AKSS2D neck · Power Tower Foreign Object | AP50:95: 71.9 -> 73.3 | Improved |
| use AKCAttention instead of SE (ablation) | YOLOv8 · Railway Pedestrian Detection | mAP: 93.9 -> 94.3 | Improved |
| use AKCAttention instead of MHA (ablation) | YOLOv8 · Railway Pedestrian Detection | FPS: 26.9 -> 29.2 | Improved |
| add all modules to C2f baseline (ablation) | YOLOv8 · Module efficiency analysis | mAP gain: +1.6% mAP for 0.8ms more latency | Improved |
| AKCMamba-YOLO vs Mamba YOLO-T on COCO | AKCMamba-YOLO · COCO2017 val | mAP: 45.4 -> 46.3 | Improved |
Closest papers by meaning
- Mamba YOLO: A Simple Baseline for Object Detection with State Space Model (AAAI 2025 · similarity 0.91)
- YOLOv10: Real-Time End-to-End Object Detection (NeurIPS 2024 · similarity 0.89)
- YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary (ICLR 2025 · similarity 0.88)
- YOLO-World: Real-Time Open-Vocabulary Object Detection (CVPR 2024 · similarity 0.88)
- YOLOv12: Attention-Centric Real-Time Object Detectors (NeurIPS 2025 · similarity 0.87)
- State Space Model Meets Transformer: A New Paradigm for 3D Object Detection (ICLR 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.
