TY - GEN
T1 - A new two-stage object detection network without RoI-Pooling
AU - Yan, Chao
AU - Chen, Weihai
AU - Chen, Peter C.Y.
AU - Kendrick, Amezquita S.
AU - Wu, Xingming
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/7/6
Y1 - 2018/7/6
N2 - Two-stage object detection networks often propose a set of candidate boxes in the first stage, and then fine- tune the boxes in the second stage. The original two-stage object detection methods mostly process the features among the candidate boxes in the picture by RoI-Pooling [3]. Due to the overlaps of the candidate boxes proposed in the first stage, the calculation of the second stage is repetitive and the single-frame detection is slow. RoI-Pooling also makes the features of the elongated shape deformed. In this paper, we present a new two-step object detection network, called Spatial Alignment Network(SAN), which does not use the RoI-Pooling layer and reduces the computational repeatability of the second stage. We also use atrous convolution for the network fine-tuning. Our network has a competitive result, and faster than the original two-stage detectors.
AB - Two-stage object detection networks often propose a set of candidate boxes in the first stage, and then fine- tune the boxes in the second stage. The original two-stage object detection methods mostly process the features among the candidate boxes in the picture by RoI-Pooling [3]. Due to the overlaps of the candidate boxes proposed in the first stage, the calculation of the second stage is repetitive and the single-frame detection is slow. RoI-Pooling also makes the features of the elongated shape deformed. In this paper, we present a new two-step object detection network, called Spatial Alignment Network(SAN), which does not use the RoI-Pooling layer and reduces the computational repeatability of the second stage. We also use atrous convolution for the network fine-tuning. Our network has a competitive result, and faster than the original two-stage detectors.
KW - Computer Vision
KW - Deep Learning
KW - Object Detection
UR - https://www.scopus.com/pages/publications/85050867180
U2 - 10.1109/CCDC.2018.8407398
DO - 10.1109/CCDC.2018.8407398
M3 - 会议稿件
AN - SCOPUS:85050867180
T3 - Proceedings of the 30th Chinese Control and Decision Conference, CCDC 2018
SP - 1680
EP - 1685
BT - Proceedings of the 30th Chinese Control and Decision Conference, CCDC 2018
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 30th Chinese Control and Decision Conference, CCDC 2018
Y2 - 9 June 2018 through 11 June 2018
ER -