Research map / Architectures and efficiency / Spiking networks
SDTrack: A Baseline for Event-based Tracking via Spiking Neural Networks
Yimeng Shan, Zhenbang Ren, Haodi Wu, Wenjie Wei, Rui-Jie Zhu, Shuai Wang, Dehao Zhang, Yichen Xiao, Jieyuan Zhang, Kexin Shi, Jingzhinan Wang, Jason K. Eshraghian and 2 more
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 Intrinsic Position Learning (IPL) (ablation) | SDTrack-Tiny · FE108 | PR (%): 89.66 -> 91.30 | Improved |
| add extra positional encoding on top of IPL (ablation) | SDTrack-Tiny · FE108 | AUC (%): 59.00 -> 58.79 (learnable), 58.57 (sinusoidal) | Hurt |
| enlarge template-search intersection to 128 (ablation) | SDTrack-Tiny · FE108 | AUC (%): 59.00 -> 43.91 | Hurt |
| use center head instead of corner head (ablation) | SDTrack-Tiny · FE108 | AUC (%): 58.81 -> 59.00 | Improved |
| add ImageNet-1K pretraining (ablation) | SDTrack-Tiny · FE108 | AUC (%): 47.80 -> 59.00 | Improved |
| use SDTrack backbone instead of Spike-driven Transformer V1 (ablation) | SDTrack-Tiny · FE108 | AUC (%): 52.4 -> 59.0 | Improved |
| use SDTrack backbone instead of Spike-driven Transformer V3 (ablation) | SDTrack-Tiny · FE108 | AUC (%): 58.9 -> 59.0 | No clear effect |
| add spike neuron before conv in decision layer (ablation) | SDTrack-Tiny · FE108 | PR (%): 91.3 -> 90.4 | Hurt |
| scale from SDTrack-Tiny to SDTrack-Base | SDTrack · VisEvent | AUC (%): 35.6 -> 37.4 | Improved |
| use T=2x2 LIF/I-LIF neurons instead of 1x4 I-LIF | SDTrack-Tiny · FELT | AUC (%): 39.3 -> 35.7 | Hurt |
| use GTP event aggregation instead of other aggregation methods | five trackers · FE108 | AUC | Improved |
Closest papers by meaning
- Fully Spiking Neural Networks for Unified Frame-Event Object Tracking (NeurIPS 2025 · similarity 0.92)
- 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 0.90)
- Leveraging Asynchronous Spiking Neural Networks for Ultra Efficient Event-Based Visual Processing (AAAI 2025 · similarity 0.89)
- CREST: An Efficient Conjointly-trained Spike-driven Framework for Event-based Object Detection Exploiting Spatiotemporal Dynamics (AAAI 2025 · similarity 0.87)
- Hybrid Spiking Vision Transformer for Object Detection with Event Cameras (ICML 2025 · 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.
