Detecting and Counting People Faces in Images Using Deep Learning Approach

Abstract

Computer vision has received increased attention in recent years. Researchers have made remarkable advancement in the field of computer vision. Computer vision applications include object detection and image classification. Object detection is to detect objects in the images as car, dog, cat, and person. People counting based on face detection remains a challenging task and still an open problem in computer vision. This thesis introduces two approaches for detecting and counting people's faces. These approaches are based on Faster-RCNN and SSD (Liu et al.), which are models used for object detection. We use these approaches for Face Detection and counting these faces. These models are deep neural networks which are trained on object detection task. In this work, we train Faster-RCNN and SSD models on Wider-Face dataset, which is composed of faces in a variety of conditions relating to occlusion, illumination, expression, pose and scale. The evaluation result on the test part of the wider face dataset is 0.5 of accuracy for Faster-RCNN and SSD, also the MRE for the Faster-RCNN is 0.3 and the SSD is 0.4. The MAE for the Faster-RCNN is 7.5 and the SSD is 8.6.