arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.10905cs.CV
收藏

具有识别特征的视频图像联合检测与跟踪

Joint Detection and Tracking in Videos with Identification Features

OSRAM GmbH(欧司朗有限公司);Technische Universität München(慕尼黑工业大学);Sapienza University of Rome(罗马第一大学)

Bharti Munjal, Abdul Rafey Aftab, Sikandar Amin, Meltem D. Brandlmaier, Federico Tombari, Fabio Galasso

摘要:

最近的工作已经表明,在视频数据的情况下,组合目标检测和跟踪任务导致两个任务的更高性能,但是它们需要高帧速率作为对性能的严格要求。在实际应用中,当模型在嵌入式设备上运行时,通常每秒只有几帧时,这一假设经常被违反。低帧速率的视频遭受大的对象位移。这里,重新识别特征可以支持匹配大位移的目标检测,但是当前的联合检测和重新识别公式降低了检测器性能,因为这两者是对比任务。在现实世界的应用程序中,具有单独的检测器和re-id模型通常是不可行的,因为内存和运行时间都有效地加倍。为了实现适用于计算能力降低的设备的鲁棒长期跟踪,我们提出了视频的检测,跟踪和重新识别特征的第一个联合优化。值得注意的是,我们的联合优化保持了检测器的性能,这是一个典型的多任务挑战。在推理时,当对象在图像中可见、可检测且缓慢移动时,我们利用检测进行跟踪(通过检测进行跟踪)。相反,我们利用重新识别功能来匹配几帧内消失(例如由于遮挡)或由于快速运动(或低帧率视频)而未被跟踪的对象。我们提出的方法在MOT上达到了最先进的水平,在UA-DETRAC'18在线跟踪器跟踪挑战中排名第一,总体排名第三。

英文摘要:

Recent works have shown that combining object detection and tracking tasks, in the case of video data, results in higher performance for both tasks, but they require a high frame-rate as a strict requirement for performance. This is assumption is often violated in real-world applications, when models run on embedded devices, often at only a few frames per second. Videos at low frame-rate suffer from large object displacements. Here re-identification features may support to match large-displaced object detections, but current joint detection and re-identification formulations degrade the detector performance, as these two are contrasting tasks. In the real-world application having separate detector and re-id models is often not feasible, as both the memory and runtime effectively double. Towards robust long-term tracking applicable to reduced-computational-power devices, we propose the first joint optimization of detection, tracking and re-identification features for videos. Notably, our joint optimization maintains the detector performance, a typical multi-task challenge. At inference time, we leverage detections for tracking (tracking-by-detection) when the objects are visible, detectable and slowly moving in the image. We leverage instead re-identification features to match objects which disappeared (e.g. due to occlusion) for several frames or were not tracked due to fast motion (or low-frame-rate videos). Our proposed method reaches the state-of-the-art on MOT, it ranks 1st in the UA-DETRAC'18 tracking challenge among online trackers, and 3rd overall.

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。

↑