Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2509.13784 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915499134156800 |
|---|---|
| author | Liang, Hanfang Wang, Bing Zhang, Shizhen Jiang, Wen Yang, Yizhuo Guo, Weixiang Yuan, Shenghai |
| author_facet | Liang, Hanfang Wang, Bing Zhang, Shizhen Jiang, Wen Yang, Yizhuo Guo, Weixiang Yuan, Shenghai |
| contents | Event cameras capture asynchronous pixel-level brightness changes with microsecond temporal resolution, offering unique advantages for high-speed vision tasks. Existing methods often convert event streams into intermediate representations such as frames, voxel grids, or point clouds, which inevitably require predefined time windows and thus introduce window latency. Meanwhile, pointwise detection methods face computational challenges that prevent real-time efficiency due to their high computational cost. To overcome these limitations, we propose the Variable-Rate Spatial Event Mamba, a novel architecture that directly processes raw event streams without intermediate representations. Our method introduces a lightweight causal spatial neighborhood encoder to efficiently capture local geometric relations, followed by Mamba-based state space models for scalable temporal modeling with linear complexity. During inference, a controller adaptively adjusts the processing speed according to the event rate, achieving an optimal balance between window latency and inference latency. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_13784 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CETUS: Causal Event-Driven Temporal Modeling With Unified Variable-Rate Scheduling Liang, Hanfang Wang, Bing Zhang, Shizhen Jiang, Wen Yang, Yizhuo Guo, Weixiang Yuan, Shenghai Computer Vision and Pattern Recognition Event cameras capture asynchronous pixel-level brightness changes with microsecond temporal resolution, offering unique advantages for high-speed vision tasks. Existing methods often convert event streams into intermediate representations such as frames, voxel grids, or point clouds, which inevitably require predefined time windows and thus introduce window latency. Meanwhile, pointwise detection methods face computational challenges that prevent real-time efficiency due to their high computational cost. To overcome these limitations, we propose the Variable-Rate Spatial Event Mamba, a novel architecture that directly processes raw event streams without intermediate representations. Our method introduces a lightweight causal spatial neighborhood encoder to efficiently capture local geometric relations, followed by Mamba-based state space models for scalable temporal modeling with linear complexity. During inference, a controller adaptively adjusts the processing speed according to the event rate, achieving an optimal balance between window latency and inference latency. |
| title | CETUS: Causal Event-Driven Temporal Modeling With Unified Variable-Rate Scheduling |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.13784 |