Real time Person Re–Identification using Deep Association Metric with Multiple Inputs

While the person re-identification problem could be seen as a challenging task, where every person wearing a certain type of clothing is recalled from the cameras, we are more concerned with distinguishing within these categories to find an exact person. There are many different challenges to solving this problem. Most real-world cameras have an insufficient resolution or are too far away from a detected person to perform high-level feature extraction. In our method, we propose a real-time person re-identification approach with multiple inputs.