Pedestrian re-identification method based on depth multi-loss fusion model

Abstract

The invention relates to a pedestrian re-identification method based on a depth multi-loss fusion model. According to the method, a deep learning technology is used, pre-processing operations such asoverturning, cutting, random erasing and style migration are carried out on the training set picture. Feature extraction is carried out through a basic network model. A plurality of loss functions areused for carrying out fusion joint training on the network. Compared with a pedestrian re-identification algorithm based on deep learning, since the method uses a plurality of preprocessing modes, fusion of three loss functions and an effective training strategy, the pedestrian re-identification performance on the data set is greatly improved. On one hand, multiple preprocessing modes can expanda data set, the generalization capability of the model is improved, the occurrence of an overfitting condition is avoided, and on the other hand, the three loss functions have respective advantages and disadvantages, and when the three loss functions are effectively combined, the used model can obtain a better recognition result.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a pedestrian re-identification method based on a deep multi-loss fusion model. Background technique [0002] The basic task of a distributed multi-camera surveillance system is to associate people with camera views at different locations and at different times, which is called the pedestrian re-identification task. It supports many key applications such as long-duration multi-camera tracking and forensic search. In fact, each camera can shoot from different angles and distances, under different lighting conditions, occlusion degrees, and different static and dynamic backgrounds. This brings some great challenges to the person re-identification task. At the same time, pedestrian re-identification techniques relying on traditional biometrics such as facial recognition are neither feasible nor reliable because pedestrians observed by cameras at unknown distances may have congested b...

AI Technical Summary

Problems solved by technology

Similarly, the Triplet model takes three images as input to shorten the intra-class distance and widen the inter-class distance, but also does not use the annotation information of a single image

In addition, research by relevant scholars has shown that there is still a relatively large intra-class distance in the Triplet loss training network.

Images