RobustHumanDetectionUnder.PDFVIP

Robust Human Detection Under Occlusion by Integrating Face and Person Detectors William Robson Schwartz1, Raghuraman Gopalan2, Rama Chellappa2, and Larry S. Davis1 1 University of Maryland, Department of Computer Science College Park, MD, USA, 20742 2 University of Maryland, Department of Electrical and Computer Engineering College Park, MD, USA, 20742 Abstract. Human detection under occlusion is a challenging problem in computer vision. We address this problem through a framework which integrates face detection and person detection. We ?rst investigate how the response of a face detector is correlated with the response of a person detector. From these observations, we formulate hypotheses that capture the intuitive feedback between the responses of face and person detectors and use it to verify if the individual detectors’ outputs are true or false. We illustrate the performance of our integration framework on challeng- ing images that have considerable amount of occlusion, and demonstrate its advantages over individual face and person detectors. 1 Introduction Human detection (face and the whole body) in still images is of high interest in computer vision. However, it is a challenging problem due to the presence of variations in people’s poses, lighting conditions, inter- and intra- person occlu- sion, amongst others. Occlusion, in particular, poses a signi?cant challenge due to the large amount of variations it implies on the appearance of the visible parts of a person. There are many human detection algorithms in the literature. In general, they fall into two categories: subwindow-based and part-based approaches. In the former category, features extracted from subwindows located within a detection window are used to describe the whole body. Subwindow-based approaches can be based on di?erent types and