公开(公告)号：GB2605727A

公开(公告)日：2022-10-12

申请号：GB202209147

申请日：2020-10-26

Applicant: IBM

Inventor： YUBO LI ,  QING WANG ,  KE JIN ,  SHI LEI ZHANG ,  JIE ZHANG

IPC: G06T7/20 ,  G06K9/00

Abstract: Technology for performing object tracking of an object in a video image using both object detection algorithm and an object tracking algorithm. Because object detection algorithms are generally slower, but accurate, than object tracking algorithms, only certain sampled video frames are subject to the object detection algorithm. The non-sampled frames locate the object using the object tracking algorithm, using the most recent previous object detected frame as a starting frame. In this way, the object detection algorithm is used to periodically, effectively calibrate the object tracking algorithm by using a lagging detection result to hide latency introduced by the object detection algorithm.

公开(公告)号：GB2605236A

公开(公告)日：2022-09-28

申请号：GB202117081

申请日：2021-11-26

Applicant: IBM

Inventor： LINGYUN WANG ,  QING WANG ,  SHI LEI ZHANG

IPC: G06T7/73 ,  G06T7/174

Abstract: A method includes obtaining from a video having a first resolution, a set of frames having a second resolution with the first resolution being higher than the second resolution and the set of frames includes a first frame and a second frame from which difference feature map is generated. A third frame having the first resolution and a third frame area is then obtained from the video and a first location of an object is detected in the third frame using the difference feature map. A first area smaller than the third is then cropped from the third frame and a feature map generated of this first area. A spatial attention area is then generated based on the difference feature map and the first feature map and used to detect the first object in the first cropped area.