An Efficient Method of Face and Keypoint Detection Based on Shared Network
We propose a method to perform face detection and facial keypoint detection in parallel.On the basis of the traditional object detection framework Faster R-CNN for face detection,we have designed a facial keypoint detection network with the same base network.And the two networks share the weights of the base network by the method of alternate training.Depends on the feature map in Faster R-CNN,after detecting the face box,we can find the feature patch corresponding to the face box directly.In this way,we make the two models can be fused together.Compared with the traditional serial process,our model run faster more than 20%,with the same accuracy.
face detection facial keypoint detection fused network
Xiaogang Tian Xiaoye Fan Feixue Tang Xixin Cao
School of software & Microelectronics,Peking University,China
国际会议
深圳
英文
90-93
2018-01-21（万方平台首次上网日期，不代表论文的发表时间）