提出高效事件相机动作识别新框架，精度提升且参数减少三成。
SMV-EAR: Bring Spatiotemporal Multi-View Representation Learning into Efficient Event-Based Action Recognition
- 通过平移不变的密集事件转换构建时空多视角表征
- 动态双分支融合使不同视角运动特征互补，精度提升超10%
- 仿生时间扭曲增强模拟真实动作速度变化，适合低功耗场景
事件相机动作识别（EAR）具备隐私保护和高效的优势，其中时间运动动态至关重要。现有时空多视角表征学习（SMVRL）方法虽通过沿空间轴H和W投影事件展现潜力，但受限于平移敏感的空间分箱表示和简单的早期拼接融合结构。本文重新审视EAR中SMVRL的关键设计环节，提出：(i) 基于平移不变的稀疏事件密集转换的合理时空多视角表征；(ii) 双分支动态融合架构，建模不同视角间样本级运动特征的互补性；(iii) 生物启发的时间扭曲增强，模拟真实人类动作的速度变异性。在HARDVS、DailyDVS-200和THU-EACT-50-CHL三个挑战性EAR数据集上，相比现有SMVRL EOR方法，分别实现+7.0%、+10.7%和+10.2%的Top-1准确率提升，同时参数减少30.1%，计算量降低35.7%，确立了该框架作为新型高效EAR范式。
原文摘要 · Abstract (English)
Event cameras action recognition (EAR) offers compelling privacy-protecting and efficiency advantages, where temporal motion dynamics is of great importance. Existing spatiotemporal multi-view representation learning (SMVRL) methods for event-based object recognition (EOR) offer promising solutions by projecting H-W-T events along spatial axis H and W, yet are limited by its translation-variant spatial binning representation and naive early concatenation fusion architecture. This paper reexamines the key SMVRL design stages for EAR and propose: (i) a principled spatiotemporal multi-view representation through translation-invariant dense conversion of sparse events, (ii) a dual-branch, dynamic fusion architecture that models sample-wise complementarity between motion features from different views, and (iii) a bio-inspired temporal warping augmentation that mimics speed variability of real-world human actions. On three challenging EAR datasets of HARDVS, DailyDVS-200 and THU-EACT-50-CHL, we show +7.0%, +10.7%, and +10.2% Top-1 accuracy gains over existing SMVRL EOR method with surprising 30.1% reduced parameters and 35.7% lower computations, establishing our framework as a novel and powerful EAR paradigm.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。