多级调优的人脸检测网络
|
摘要
为了提高人脸检测的速度和精度,提出了一个端到端的人脸检测网络。该网络包含两个模块:粗筛模块和精调模块。粗筛模块负责过滤背景候选框,仅保留少量高质量候选框,以提高检测的速度;精调模块则负责进一步遴选剩余的候选框并对其进行调整来适应真实的人脸,以提高检测的精度。另外,选择性地对一些层的特征进行融合从而进一步提高网络的精度。在FDDB数据集上,该网络在500次假阳性误报的情况下取得了95.71%的离散得分和75.29%的连续得分。
In order to improve the speed and accuracy of face detection, an end-to-end face detection network was proposed. It includes two main modules: coarse-scale module and fine-scale module. The coarse-scale module was used to filter the background proposals and keep only a small amount of high-quality proposals to speed up the detection. The fine-scale module was used to further filter the remaining proposals and adjust the selected proposals to improve the detection accuracy. Besides, features of some layers were selected to fuse together in order to further improve sccuracy of the network. On FDDB dataset, the proposed network achieves 95.71% of discrete score and 75.29% of continuous score with 500 false positives.
In order to improve the speed and accuracy of face detection, an end-to-end face detection network was proposed. It includes two main modules: coarse-scale module and fine-scale module. The coarse-scale module was used to filter the background proposals and keep only a small amount of high-quality proposals to speed up the detection. The fine-scale module was used to further filter the remaining proposals and adjust the selected proposals to improve the detection accuracy. Besides, features of some layers were selected to fuse together in order to further improve sccuracy of the network. On FDDB dataset, the proposed network achieves 95.71% of discrete score and 75.29% of continuous score with 500 false positives.
引文
[1] VIOLA P,JONES M J.Robust real-time face detection [J].International Journal of Computer Vision,2004,57(2):137-154.
[2] 苑玮琦,韩春霞.复杂背景下的多人脸检测方法 [J].计算机应用,2010,30(3):635-638.
[3] 娄康,薛彦兵,张桦,等.基于代价敏感深度决策树的公交车环境人脸检测 [J].计算机应用,2017,37(11):3152-3156.
[4] REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149.
[5] JIANG H,LEARNED-MILLER E.Face detection with the Faster R-CNN [C]// Proceedings of the 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition.Piscataway,NJ:IEEE,2017:650-657.
[6] SHAO W,WANG H,ZHENG Y,et al.Compact face representation via forward model selection[C]// Proceedings of the 2016 Chinese Conference on Biometric Recognition.Berlin:Springer,2016:112-120.
[7] YE H,SHAO W Y,WANG H,et al.Face recognition via active annotation and learning [C]// Proceedings of the 24th ACM Conference on Multimedia.New York:ACM,2016:1058-1062.
[8] RANJAN R,PATEL V M,CHELLAPPA R.HyperFace:a deep multi-task learning framework for face detection,landmark localization,pose estimation,and gender recognition [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2019,41(1):121-135.
[9] LI Y,SUN B,WU T,et al.Face detection with end-to-end integration of a convnet and a 3D model[C]// Proceedings of the 2016 European Conference on Computer Vision.Berlin:Springer,2016:420-436.
[10] YU J,JIANG Y,WANG Z,et al.UNITBOX:an advanced object detection network [C]// Proceedings of the 24th ACM International Conference on Multimedia.New York:ACM,2016:516-520.
[11] MA J,SHAO W,YE H,et al.Arbitrary-oriented scene text detection via rotation proposals [J].IEEE Transactions on Multimedia,2018,20(11):3111-3122.
[12] JAIN V,LEARNED-MILLER E G.FDDB:a benchmark for face detection in unconstrained settings,UM-CS-2010-009[R].Amherst:University of Massachusetts,2010.
[13] ZHU X,RAMAN D.Face detection,pose estimation,and landmark localization in the wild[C]// Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2012:2879-2886.
[14] YANG S,LUO P,LOY C C,et al.WIDER FACE:A face detection benchmark [C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2016:5525-5533.
[15] YANG S,LUO P,LOY C C,et al.From facial parts responses to face detection:A deep learning approach [C]// Proceedings of the 2015 International Conference on Computer Vision.Washington,DC:IEEE Computer Society,2015:3676-3684.
[16] MATHIAS M,BENENSON R,PEDERSOLI M,et al.Face detection without 440 bells and whistles [C]// Proceedings of the 2014 European Conference on Computer Vision.Berlin:Springer,2014:720-735.
[17] YAN J,ZHANG X,LEI Z,et al.Face detection by structural models [J].Image and Vision Computing,2015,32(10):790-799.
[18] SHEN X,LIN Z,BRANDT J,et al.Detecting and aligning faces by image retrieval [C]// Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2013:3460-3467.
[2] 苑玮琦,韩春霞.复杂背景下的多人脸检测方法 [J].计算机应用,2010,30(3):635-638.
[3] 娄康,薛彦兵,张桦,等.基于代价敏感深度决策树的公交车环境人脸检测 [J].计算机应用,2017,37(11):3152-3156.
[4] REN S,HE K,GIRSHICK R,et al.Faster R-CNN:Towards real-time object detection with region proposal networks [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(6):1137-1149.
[5] JIANG H,LEARNED-MILLER E.Face detection with the Faster R-CNN [C]// Proceedings of the 2017 12th IEEE International Conference on Automatic Face & Gesture Recognition.Piscataway,NJ:IEEE,2017:650-657.
[6] SHAO W,WANG H,ZHENG Y,et al.Compact face representation via forward model selection[C]// Proceedings of the 2016 Chinese Conference on Biometric Recognition.Berlin:Springer,2016:112-120.
[7] YE H,SHAO W Y,WANG H,et al.Face recognition via active annotation and learning [C]// Proceedings of the 24th ACM Conference on Multimedia.New York:ACM,2016:1058-1062.
[8] RANJAN R,PATEL V M,CHELLAPPA R.HyperFace:a deep multi-task learning framework for face detection,landmark localization,pose estimation,and gender recognition [J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2019,41(1):121-135.
[9] LI Y,SUN B,WU T,et al.Face detection with end-to-end integration of a convnet and a 3D model[C]// Proceedings of the 2016 European Conference on Computer Vision.Berlin:Springer,2016:420-436.
[10] YU J,JIANG Y,WANG Z,et al.UNITBOX:an advanced object detection network [C]// Proceedings of the 24th ACM International Conference on Multimedia.New York:ACM,2016:516-520.
[11] MA J,SHAO W,YE H,et al.Arbitrary-oriented scene text detection via rotation proposals [J].IEEE Transactions on Multimedia,2018,20(11):3111-3122.
[12] JAIN V,LEARNED-MILLER E G.FDDB:a benchmark for face detection in unconstrained settings,UM-CS-2010-009[R].Amherst:University of Massachusetts,2010.
[13] ZHU X,RAMAN D.Face detection,pose estimation,and landmark localization in the wild[C]// Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2012:2879-2886.
[14] YANG S,LUO P,LOY C C,et al.WIDER FACE:A face detection benchmark [C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2016:5525-5533.
[15] YANG S,LUO P,LOY C C,et al.From facial parts responses to face detection:A deep learning approach [C]// Proceedings of the 2015 International Conference on Computer Vision.Washington,DC:IEEE Computer Society,2015:3676-3684.
[16] MATHIAS M,BENENSON R,PEDERSOLI M,et al.Face detection without 440 bells and whistles [C]// Proceedings of the 2014 European Conference on Computer Vision.Berlin:Springer,2014:720-735.
[17] YAN J,ZHANG X,LEI Z,et al.Face detection by structural models [J].Image and Vision Computing,2015,32(10):790-799.
[18] SHEN X,LIN Z,BRANDT J,et al.Detecting and aligning faces by image retrieval [C]// Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition.Washington,DC:IEEE Computer Society,2013:3460-3467.