arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1807.05073cs.CV
收藏

基于3D卷积网络和非局部注意力的视频人物再识别

Video-based Person Re-identification via 3D Convolutional Networks and Non-local Attention

University of Science and Technology of China(中国科学技术大学);University of Chinese Academy of Sciences(中国科学院大学);Megvii Inc.(Face++)(旷视科技(Face++))

Xingyu Liao, Lingxiao He, Zhouwang Yang, Chi Zhang

摘要:

基于视频的人物重新识别(ReID)是一个具有挑战性的问题,其中跨非重叠摄像机的人物的一些视频轨迹可用于匹配。从视频轨迹进行特征聚合是基于视频的人ReID的关键步骤。许多现有的方法通过平均/最大时间池或RNN来解决这个问题。然而,这些方法不能同时处理时间依赖性和空间错位问题。我们受到视频动作识别的启发,它涉及从视频轨迹中识别不同的动作。首先,我们使用3D卷积的视频体积,而不是使用跨帧的2D卷积,同时提取空间和时间特征。其次,我们使用一个非局部块来解决错位问题,并捕捉时空长范围的依赖关系。因此,网络可以学习有用的时空信息,作为输入特征图中所有空间和时间位置的特征的加权和。三个数据集上的实验结果表明,我们的框架优于国家的最先进的方法,在多个指标上有很大的利润。

英文摘要:

Video-based person re-identification (ReID) is a challenging problem, where some video tracks of people across non-overlapping cameras are available for matching. Feature aggregation from a video track is a key step for video-based person ReID. Many existing methods tackle this problem by average/maximum temporal pooling or RNNs with attention. However, these methods cannot deal with temporal dependency and spatial misalignment problems at the same time. We are inspired by video action recognition that involves the identification of different actions from video tracks. Firstly, we use 3D convolutions on video volume, instead of using 2D convolutions across frames, to extract spatial and temporal features simultaneously. Secondly, we use a non-local block to tackle the misalignment problem and capture spatial-temporal long-range dependencies. As a result, the network can learn useful spatial-temporal information as a weighted sum of the features in all space and temporal positions in the input feature map. Experimental results on three datasets show that our framework outperforms state-of-the-art approaches by a large margin on multiple metrics.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑