TY - JOUR
T1 - Auxiliary Detection Head for One-Stage Object Detection
AU - Jin, Guozheng
AU - Taniguchi, Rin Ichiro
AU - Qu, Fengzhong
N1 - Funding Information:
This work was supported in part by the project of China Scholarship Council under Grant 201806320324, in part by the National Natural Science Foundation of China for Excellent Young Scholars under Grant 61722113, in part by the Harbin Engineering University Open Project under Grant SSJSWDZC2018016, and in part by the Joint Fund of National Natural Science Foundation of China and Zhejiang Province under Grant U1809211.
Publisher Copyright:
© 2013 IEEE.
PY - 2020
Y1 - 2020
N2 - The auxiliary classifier can improve the performance of classification networks. However, the utility of the auxiliary detection head has not been explored in the object detection field. In this paper, we propose an auxiliary detection head to boost the performance of one-stage object detectors. Similar to other detection heads, the auxiliary detection head consists of a classification subnet and a regression subnet, which are essentially two convolution layers. Thus the auxiliary detection head is computationally efficient. Besides, the auxiliary detection head achieves implicit two-step cascaded regression. Specifically, the auxiliary detection head uses its output boxes as anchors for further regression. Within the auxiliary detection head, refinement of object localization corresponds to adjust the positions of its output boxes towards ground truth boxes, which helps the network learn more robust features. At inference, the auxiliary detection head can be removed without any adverse effect on the performance of the main detector head, which benefits from its independence and leads to two advantages: shrink the model size and shorten inference time. The proposed method is evaluated on Pascal VOC and COCO datasets. By incorporating the auxiliary detection head into a state-of-the-art object detector in parallel with the main detection head, we show consistent improvement over its performance on different benchmarks, whereas no extra parameters are introduced at inference time.
AB - The auxiliary classifier can improve the performance of classification networks. However, the utility of the auxiliary detection head has not been explored in the object detection field. In this paper, we propose an auxiliary detection head to boost the performance of one-stage object detectors. Similar to other detection heads, the auxiliary detection head consists of a classification subnet and a regression subnet, which are essentially two convolution layers. Thus the auxiliary detection head is computationally efficient. Besides, the auxiliary detection head achieves implicit two-step cascaded regression. Specifically, the auxiliary detection head uses its output boxes as anchors for further regression. Within the auxiliary detection head, refinement of object localization corresponds to adjust the positions of its output boxes towards ground truth boxes, which helps the network learn more robust features. At inference, the auxiliary detection head can be removed without any adverse effect on the performance of the main detector head, which benefits from its independence and leads to two advantages: shrink the model size and shorten inference time. The proposed method is evaluated on Pascal VOC and COCO datasets. By incorporating the auxiliary detection head into a state-of-the-art object detector in parallel with the main detection head, we show consistent improvement over its performance on different benchmarks, whereas no extra parameters are introduced at inference time.
UR - http://www.scopus.com/inward/record.url?scp=85085086194&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85085086194&partnerID=8YFLogxK
U2 - 10.1109/ACCESS.2020.2992532
DO - 10.1109/ACCESS.2020.2992532
M3 - Article
AN - SCOPUS:85085086194
VL - 8
SP - 85740
EP - 85749
JO - IEEE Access
JF - IEEE Access
SN - 2169-3536
M1 - 9086609
ER -