Computer Vision

Deep neural network-based object detection has become the cornerstone of many real-world applications. Along with this success comes concerns about its vulnerability to malicious attacks. To gain more insight into this issue, we proposed a contextual camouflage attack (CCA for short) algorithm to influence the performance of object detectors.

Person re-identification(re-id) has been widespread concerned recently, as this issue underpins various critical applications such as video surveillance, pedestrian tracking, and searching. In this project, we proposed a novel video-based person re-id framework using multiple Convolutional Neural Networks.