Research map / Detection and segmentation / CNN features
EAS-SNN: End-to-End Adaptive Sampling and Representation for Event-based Detection with Recurrent Spiking Neural Networks
Ziming Wang, Ziling Wang, Huaning Li, Lang Qin, Runhao Jiang, De Ma, Huajin Tang
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 recurrent synaptic connections to SNN sampler (ablation) | SNN potential aggregation · N-Caltech 101 | mAP50: 0.567 -> 0.634 | Improved |
| add recurrent synaptic connections to SNN sampler (ablation) | SNN potential aggregation · Gen1 | mAP50: 0.670 -> 0.719 | Improved |
| add adaptive spike-based sampling without SAT or RPD (ablation) | RSNN · N-Caltech 101 | mAP50: 0.634 -> 0.600 | Hurt |
| add RPD or SAT alone to adaptive sampling (ablation) | ARSNN · N-Caltech 101 | mAP50: 0.634 (RSNN) -> 0.586 (RPD only), 0.592 (SAT only) | Hurt |
| combine SAT and RPD in ARSNN (ablation) | ARSNN · N-Caltech 101 | mAP50: 0.634 -> 0.664 | Improved |
| combine SAT and RPD in ARSNN (ablation) | ARSNN · Gen1 | mAP50: 0.719 -> 0.731 | Improved |
| use learned spike-based sampler instead of Voxel Grid, Time Surface or Voxel Cube | event representation · Gen1 | mAP50: 0.654 (Time Surface) -> 0.731 | Improved |
| use spiking FPN instead of non-spiking FPN in compact network (ablation) | EAS-SNN (S) · Gen1 | mAP50:95: spiking FPN 0.372 vs fully spiking 0.354 | Mixed |
| use spiking detection head within few time steps (ablation) | EAS-SNN · Gen1 | mAP50:95: 0.372 -> 0.354 (S); 0.393 -> 0.375 (M, spiking FPN) | Hurt |
| replace event-driven GPU spiking detector with spike-based detector over ANN | YOLOX · Gen1 | energy: 164.35 mJ -> 28.10 mJ (YOLOX-M) | Improved |
| EAS-SNN vs best spike-based baseline | Gen1 detection · Gen1 | mAP50:95: 0.310 -> 0.354 | Improved |
Later papers that compared with this method
2 later accepted papers report EAS-SNN as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Bio-Vision-Inspired Spiking Neural Networks for Object Detection with Event Cameras (ICML 2026 · better on 3 of 3 shared results)
- Object Detection using Event Camera: A MoE Heat Conduction based Detector and A New Benchmark Dataset (CVPR 2025 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Efficient Event-Based Object Detection: A Hybrid Neural Network with Spatial and Temporal Attention (CVPR 2025 · similarity 0.92)
- SFOD: Spiking Fusion Object Detector (CVPR 2024 · similarity 0.91)
- CREST: An Efficient Conjointly-trained Spike-driven Framework for Event-based Object Detection Exploiting Spatiotemporal Dynamics (AAAI 2025 · similarity 0.91)
- SpikePoint: An Efficient Point-based Spiking Neural Network for Event Cameras Action Recognition (ICLR 2024 · similarity 0.90)
- Spiking Neural Network as Adaptive Event Stream Slicer (NeurIPS 2024 · similarity 0.88)
- Hybrid Spiking Vision Transformer for Object Detection with Event Cameras (ICML 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.
