• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of visual communication & image representation >Occlusion-robust online multi-object visual tracking using a GM-PHD filter with CNN-based re-identification
【24h】

Occlusion-robust online multi-object visual tracking using a GM-PHD filter with CNN-based re-identification

机译：Occlusion-robust online multi-object visual tracking using a GM-PHD filter with CNN-based re-identification

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相关主题

摘要

We propose a novel online multi-object visual tracker using a Gaussian mixture Probability Hypothesis Density(GM-PHD) filter and deep appearance learning. The GM-PHD filter has a linear complexity with the numberof objects and observations while estimating the states and cardinality of time-varying number of objects,however, it is susceptible to miss-detections and does not include the identity of objects. We use visual-spatiotemporalinformation obtained from object bounding boxes and deeply learned appearance representations toperform estimates-to-tracks data association for target labeling as well as formulate an augmented likelihoodand then integrate into the update step of the GM-PHD filter. We also employ additional unassigned tracksprediction after the data association step to overcome the susceptibility of the GM-PHD filter towards missdetectionscaused by occlusion. Extensive evaluations on MOT16, MOT17 and HiEve benchmark data sets showthat our tracker significantly outperforms several state-of-the-art trackers in terms of tracking accuracy andidentification.

著录项

获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号