Light-Net: Lightweight Object Detector

IEEE Access

Qifei Wang¹  Jianping Gou²  Qiujie Wang³  Junmin Xue⁴  Guo Yi⁴  Yunbo Rao⁴  Jiansu Pu⁴ 

[1] Department of Electrical Engineering and Computer Sciences (EECS), University of California at Berkeley, Berkeley, CA, USA;School of Computer Science and Telecommunication Engineering, Jiangsu University, Zhenjiang, China;School of Computer Science, Guangdong University of Technology, Guangzhou, China;School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu, China;

关键词: Object detection;    light-net;    prior box;    anchor;    object detector;   

DOI  :  10.1109/ACCESS.2020.3029592

来源: DOAJ

【 摘 要 】

Currently, object detectors based on CNN, such as RetinaNet, Faster-RCNN, CornerNet series, can achieve good performance, but have some common drawbacks, like large calculation cost, high model complexity and slow detection speed. In this paper, a new lightweight object detector is proposed, which adopted a density-based approach to merge the real boxes. To reduce calculation cost and improve detection speed, the tactic of multi-scale output is adopted to predict objects of different sizes with features of different scales. Furthermore, a new lightweight network model is proposed, which can show better performance in computation, FPS, and model complexity. Meanwhile, the separation of convolution is used to improve the basic convolution layer, which can achieve better results under the same number of filters. In the experiments, we verified the capability of our methods based on ablation experiment and model evaluation, which demonstrates the superiority of our method. Moreover, we have also conducted deep network and multichannel experiments on MS-COCO2014 datasets and achieved 20.9% mAP performance.

【 授权许可】

Unknown