研究人员开发了 SpikeTAD，一种用于视频端到端时序动作检测的新型脉冲神经网络 (SNN) 架构。该方法旨在解决传统人工神经网络的高功耗和模型尺寸大的问题，使其适用于在移动设备和神经形态芯片上部署。SpikeTAD 展现出具有竞争力的性能，在 THUMOS14 上达到 67.2% 的 mAP，在 ActivityNet-1.3 上达到 37.42% 的 mAP，同时保持显著更低的功耗。 AI
影响 为边缘设备实现更节能的视频理解模型。
排序理由 该集群包含一篇详细介绍新模型架构和基准测试结果的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2606.12033v1 Announce Type: new Abstract: Video understanding is a crucial part of computer vision, with numerous application scenarios. With the increasing popularity of mobile devices, an increasing number of efforts are trying to deploy video understanding models on them…
Video understanding is a crucial part of computer vision, with numerous application scenarios. With the increasing popularity of mobile devices, an increasing number of efforts are trying to deploy video understanding models on them. However, existing video understanding models a…