Research map / Point clouds / Point cloud classification
SpikeNet: Sparse Spike-Driven Mask Vector Transformer for Energy-Efficient and Stable Spiking Point Cloud Processing
Zhiming Zhou, Yong He, Chaoxu Mu, Qiaoyun Wu, 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 |
|---|---|---|---|
| use DSSR before AMP and SVMT after AMP instead of the reverse order (ablation) | SpikeNet · ModelNet40 | OA (%): 92.7 -> 93.4 | Improved |
| add SVMT to a DSSR-only encoder (second position) (ablation) | SpikeNet · ModelNet40 | OA (%): 92.2 -> 93.4 | Improved |
| add DSSR to an SVMT-only encoder (first position) (ablation) | SpikeNet · ModelNet40 | OA (%): 92.0 -> 93.4 | Improved |
| replace plain spiking MLP blocks with DSSR and SVMT (ablation) | SpikeNet · ModelNet40 | OA (%): 90.1 -> 93.4 | Improved |
| spiking DSSR+SVMT versus ANN Conv-BN-ReLU blocks (ablation) | SpikeNet · ModelNet40 | OA (%): 92.7 -> 93.4 | Improved |
| use Channel Masker instead of Npoint Masker in SPM (ablation) | SpikeNet · ModelNet40 | OA (%): 93.2 -> 93.4 | Improved |
| use subtraction instead of summation as attention operator tau (ablation) | SpikeNet-C · ModelNet40 | OA (%): 93.2 -> 93.4 | Improved |
| use subtraction instead of division as attention operator tau (ablation) | SpikeNet-C · ModelNet40 | OA (%): 92.7 -> 93.4 | Improved |
| use subtraction instead of Hadamard product as attention operator tau (ablation) | SpikeNet-C · ModelNet40 | OA (%): 93.0 -> 93.4 | Improved |
| increase time steps from 1 to 3 (ablation) | SpikeNet-C · ModelNet40 | OA (%): 92.8 -> 93.4 | Improved |
| increase time steps from 3 to 4 (ablation) | SpikeNet-C · ModelNet40 | OA (%): 93.4 -> 93.2 | Hurt |
| increase embedding dimension from 64 to 128 (ablation) | SpikeNet-C · ModelNet40 | OA (%): 93.4 -> 92.9 | Hurt |
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 0.91)
- Efficient 3D Recognition with Event-driven Spike Sparse Convolution (AAAI 2025 · similarity 0.91)
- Spike2Former: Efficient Spiking Transformer for High-performance Image Segmentation (AAAI 2025 · similarity 0.90)
- Noise-Injected Spiking Graph Convolution for Energy-Efficient 3D Point Cloud Denoising (AAAI 2025 · similarity 0.89)
- STEP: A Unified Spiking Transformer Evaluation Platform for Fair and Reproducible Benchmarking (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.
