Research map / Point clouds / Point cloud classification
3DSMT: A Hybrid Spiking Mamba-Transformer for Point Cloud Analysis
Zhouzhiming, Yong He, Qiaoyun Wu, Chaoxu Mu, Ajmal Saeed Mian
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 Transformer branch to spiking base (ablation) | 3DSMT SNN · ModelNet40 | OA (%): 92.1 -> 93.8 | Improved |
| add bidirectional Mamba to spiking base (ablation) | 3DSMT SNN · ModelNet40 | OA (%): 92.1 -> 94.0 | Improved |
| combine Transformer and Mamba (hybrid) (ablation) | 3DSMT · ModelNet40 | OA (%): 94.0 -> 94.7 | Improved |
| use bidirectional instead of unidirectional SSM (ablation) | 3DSMT · ModelNet40 | OA (%): 94.2 -> 94.7 | Improved |
| use IF instead of LIF neurons (ablation) | 3DSMT · ModelNet40 | OA (%): 93.8 -> 94.7 | Improved |
| use no ordering instead of Z-order or shuffle (ablation) | 3DSMT · ModelNet40 | OA (%): 93.8 -> 94.7 | Improved |
| use L-SSM + C-SSM bidirectional scan instead of L-SSM + T-SSM (ablation) | 3DSMT · ModelNet40 | OA (%): 94.3 -> 94.7 | Improved |
| set SLOA neighbors k=4 instead of 6 (ablation) | 3DSMT · ModelNet40 | OA (%): 94.0 -> 94.7 | Improved |
| use 128 tokens instead of 256 (ablation) | 3DSMT · ModelNet40 | OA (%): 94.0 -> 94.7 | Improved |
| use 12 SHB blocks instead of 16 (ablation) | 3DSMT · ModelNet40 | OA (%): 93.9 -> 94.7 | Improved |
| use 1024 points instead of 4096 (ablation) | 3DSMT · ModelNet40 | OA (%): 93.8 -> 94.7 | Improved |
| replace spiking neurons with ReLU (ablation) | 3DSMT · ModelNet40 | OA (%): 94.7 -> 94.9 | No clear effect |
Closest papers by meaning
- Spiking Point Transformer for Point Cloud Classification (AAAI 2025 · similarity 0.92)
- Point-to-Spike Residual Learning for Energy-Efficient 3D Point Cloud Classification (AAAI 2024 · similarity 0.89)
- PointMamba: A Simple State Space Model for Point Cloud Analysis (NeurIPS 2024 · similarity 0.89)
- Efficient 3D Recognition with Event-driven Spike Sparse Convolution (AAAI 2025 · similarity 0.89)
- Point Cloud Mamba: Point Cloud Learning via State Space Model (AAAI 2025 · similarity 0.88)
- Noise-Injected Spiking Graph Convolution for Energy-Efficient 3D Point Cloud Denoising (AAAI 2025 · similarity 0.88)
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.
