TY - GEN
T1 - Joint Detection, Re-Identification, and Lstm in Multi-Object Tracking
AU - Tsai, Wen-Jiin
AU - Huang, Zih Jie
AU - Chung, Chen En
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/7
Y1 - 2020/7
N2 - Using Convolutional Neural Networks (CNN) in object tracking typically utilizes spatial features, while ignores the temporal correlation of frames in the whole film, causing that it is easy to lose the target when it is occluded by other objects. To cope with the problem, a robust system combining CNN and long short-term memory (LSTM) is proposed for multi-object tracking. The system consists of three modules: object detection, data association, and LSTM tracking. With the proposed approach, the tracking accuracy can be greatly improved especially when the tracking targets suffer from occlusion. Experimental results showed that the proposed system exhibits outstanding tracking accuracy and stability.
AB - Using Convolutional Neural Networks (CNN) in object tracking typically utilizes spatial features, while ignores the temporal correlation of frames in the whole film, causing that it is easy to lose the target when it is occluded by other objects. To cope with the problem, a robust system combining CNN and long short-term memory (LSTM) is proposed for multi-object tracking. The system consists of three modules: object detection, data association, and LSTM tracking. With the proposed approach, the tracking accuracy can be greatly improved especially when the tracking targets suffer from occlusion. Experimental results showed that the proposed system exhibits outstanding tracking accuracy and stability.
KW - Convolutional neural network
KW - Long short-term memory (LSTM)
KW - Multiple object tracking (MOT)
UR - http://www.scopus.com/inward/record.url?scp=85090395251&partnerID=8YFLogxK
U2 - 10.1109/ICME46284.2020.9102884
DO - 10.1109/ICME46284.2020.9102884
M3 - Conference contribution
AN - SCOPUS:85090395251
T3 - Proceedings - IEEE International Conference on Multimedia and Expo
BT - 2020 IEEE International Conference on Multimedia and Expo, ICME 2020
PB - IEEE Computer Society
T2 - 2020 IEEE International Conference on Multimedia and Expo, ICME 2020
Y2 - 6 July 2020 through 10 July 2020
ER -