Improved pedestrian re-identification algorithm based on person attribute extraction
Pedestrian re-identification(ReID)refers to the use of computer vision techniques to determine whether specific pedestrians exist in image or video sequences.Currently,pedestrian re-identification has achieved good experimental results in theory,but there are still many challenges in engineering applications.To address the complexity of pedestrian images in practical scenarios(such as differences between cameras,influences from clothing,occlusions,poses,and viewpoints),as well as the low rec-ognition rate of pedestrians,combining pedestrian re-identification algorithms with person attribute extraction algorithms and intro-ducing attention mechanisms has improved the re-identification rate of pedestrians in engineering applications.
万方数据
维普
