Clique: Spatiotemporal Object Re-identification at the City Scale

arxiv(2020)

引用 0|浏览45
暂无评分
摘要
Object re-identification (ReID) is a key application of city-scale cameras. While classic ReID tasks are often considered as image retrieval, we treat them as spatiotemporal queries for locations and times in which the target object appeared. Spatiotemporal reID is challenged by the accuracy limitation in computer vision algorithms and the colossal videos from city cameras. We present Clique, a practical ReID engine that builds upon two new techniques: (1) Clique assesses target occurrences by clustering fuzzy object features extracted by ReID algorithms, with each cluster representing the general impression of a distinct object to be matched against the input; (2) to search in videos, Clique samples cameras to maximize the spatiotemporal coverage and incrementally adds cameras for processing on demand. Through evaluation on 25 hours of videos from 25 cameras, Clique reached a high accuracy of 0.87 (recall at 5) across 70 queries and runs at 830x of video realtime in achieving high accuracy.
更多
查看译文
关键词
city scale,object,re-identification
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要