Streaming video clips with large-scale video tokens impede vision transformers (ViTs) for efficient recognition, especially in video action detection where sufficient spatiotemporal representations are required for precise actor identification. In this work, we propose an end-to-end framework for efficient video action detection (EVAD) based on vanilla ViTs. Our EVAD consists of two specialized designs for video action detection. First, we propose a spatiotemporal token dropout from a keyframe-centric perspective. In a video clip, we maintain all tokens from its keyframe, preserve tokens relevant to actor motions from other frames, and drop out the remaining tokens in this clip. Second, we refine scene context by leveraging remaining tokens for better recognizing actor identities. The region of interest (RoI) in our action detector is expanded into temporal domain. The captured spatiotemporal actor identity representations are refined via scene context in a decoder with the attention mechanism. These two designs make our EVAD efficient while maintaining accuracy, which is validated on three benchmark datasets (i.e., AVA, UCF101-24, JHMDB). Compared to the vanilla ViT backbone, our EVAD reduces the overall GFLOPs by 43% and improves real-time inference speed by 40% with no performance degradation. Moreover, even at similar computational costs, our EVAD can improve the performance by 1.0 mAP with higher resolution inputs. Code is available at https://github.com/MCG-NJU/EVAD.

翻译：大规模视频片段中的海量视频Token阻碍了视觉Transformer（ViT）的高效识别，特别是在需要充分时空表征以精确识别动作主体的视频动作检测任务中。本文基于原始ViT架构提出了一种端到端的高效视频动作检测框架（EVAD）。我们的EVAD包含两项针对视频动作检测的专门设计。首先，我们从关键帧为中心的角度提出了一种时空Token丢弃策略：在视频片段中，保留关键帧的所有Token，保留其他帧中与动作主体运动相关的Token，并丢弃该片段中的其余Token。其次，我们利用剩余Token精炼场景上下文以更好地识别动作主体身份。动作检测器中的感兴趣区域（RoI）被扩展至时域，通过解码器中的注意力机制利用场景上下文精炼所捕获的时空动作主体身份表征。这两项设计使得EVAD在保持精度的同时实现高效，并在三个基准数据集（即AVA、UCF101-24、JHMDB）上得到验证。与原始ViT骨干网络相比，我们的EVAD在无性能损失的情况下将整体GFLOPs降低43%，实时推理速度提升40%。此外，即使在相似计算成本下，采用更高分辨率输入时EVAD仍可将性能提升1.0 mAP。代码已开源至https://github.com/MCG-NJU/EVAD。