Research map / Point clouds / Point cloud classification
Rethinking Efficient and Effective Point-Based Networks for Event Camera Classification and Regression
Hongwei Ren, Yue Zhou, Jiadong Zhu, Xiaopeng Lin, Haotian Fu, Yulong Huang, Yuetong Fang, Fei Ma, Hao Yu, Bojun Cheng
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 Mamba temporal module (vs spatial aggregation) (ablation) | EventMamba · IJRR CPR (average) | position/orientation error: 0.018m,1.239 -> 0.012m,0.831 | Improved |
| add Mamba global extractor (ablation) | EventMamba · HMDB51-DVS | accuracy (%): 57.5 -> 82.1 | Improved |
| use Mamba instead of LSTM (ablation) | EventMamba · HMDB51-DVS | accuracy (%): 70.8 -> 82.1 | Improved |
| use Mamba instead of LSTM (ablation) | EventMamba · IJRR CPR (average) | orientation error (deg): 0.904 -> 0.831 | Improved |
| use Mamba instead of LSTM (ablation) | EventMamba · IJRR CPR (average) | inference time: 6.8ms -> 5.5 ms | Improved |
| increase input points 512 to 1024 (ablation) | EventMamba · HMDB51-DVS | accuracy (%): 67.9 -> 76.6 | Improved |
| increase input points 1024 to 2048 (ablation) | EventMamba · HMDB51-DVS | accuracy (%): 76.6 -> 82.1 | Improved |
| use EventMamba instead of TTPOINT | EventMamba · UCF101-DVS | accuracy: 0.725 -> 0.979 | Improved |
| use EventMamba instead of TTPOINT | EventMamba · HMDB51-DVS | accuracy: 0.569 -> 0.864 | Improved |
| use EventMamba instead of TTPOINT | EventMamba · DVS Action | accuracy: 0.927 -> 0.891 | Hurt |
| use EventMamba instead of TTPOINT | EventMamba · Daily DVS | accuracy: 0.991 -> 0.991 | No clear effect |
| use point sampling instead of event stacking preprocessing | EventMamba · IJRR preprocessing time | time: 0.457 ms -> 0.065 ms | Improved |
Closest papers by meaning
- Scalable Event Cloud Network for Event-based Classification (ICML 2026 · similarity 0.90)
- A Simple and Effective Point-based Network for Event Camera 6-DOFs Pose Relocalization (CVPR 2024 · similarity 0.89)
- PASS: Path-selective State Space Model for Event-based Recognition (NeurIPS 2025 · similarity 0.88)
- EventMamba: Enhancing Spatio-Temporal Locality with State Space Models for Event-Based Video Reconstruction (AAAI 2025 · similarity 0.88)
- SpikePoint: An Efficient Point-based Spiking Neural Network for Event Cameras Action Recognition (ICLR 2024 · similarity 0.87)
- 3DInAction: Understanding Human Actions in 3D Point Clouds (CVPR 2024 · similarity 0.87)
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.
