A pedestrian re-recognition method based on transfer learning and feature fusion

Abstract

The invention discloses a pedestrian re-identification method based on transfer learning and feature fusion, comprising the following steps of acquiring the pedestrian data, carrying out initial training through a neural network, modifying the structure and combining the improved loss function to carry out re-training in a data set; carrying out the manual feature extraction and neural network feature extraction; after extracting features, fusing the two features to get high-Low grade features; using an XQDA algorithm to calculate the high-Low grade features, classifying and verifying to obtain the re-recognition results. The method of the invention adopts the cross entropy loss function and the triple loss function to restrict the whole network more strongly, and then extracts manual features and convolution network features to carry out feature fusion, and forms high-low features, covers different levels of pedestrian feature expression, achieves better recognition results, and reduces the training time in the form of fine-tuning, and has good generalization and portability for small data sets.

Description

technical field [0001] The invention relates to the research field of computer vision, in particular to a pedestrian re-identification method based on transfer learning and feature fusion. Background technique [0002] The person re-identification process mainly consists of two parts, namely feature learning part and metric learning part. Most of the existing person re-identification methods try to extract a good enough feature expression so that it can describe a person's unique and robust features under different conditions. For this purpose, researchers have studied color, texture and A variety of manual features are designed from different angles such as shapes, and good results have been achieved. In terms of metric learning, the use of standard distance metrics such as Euclidean distance cannot calculate the similarity of pedestrian features well, so the research on metric learning is to learn a suitable metric so that the distance between different images of the same...

AI Technical Summary

Problems solved by technology

[0004] The existing technology generally uses two methods in feature extraction. One is to construct targeted features according to the characteristics of pedestrian data, such as using color statistical features, texture features, local feature fusion (ELF), and local maximum feature (LOMO) etc., the second is to use the deep learning method to learn the implicit features of pedestrians through the neural network, such as Wu et al. designed a convolutional network structure called "PersonNet"; Liu et al. proposed a multi-scale triplet convolutional structure; Wang et al. designed a Siamese convolutional network framework; Varior et al. used a long-short-term memory network (LSTM) combined with a Siamese network, etc.; however, manual design features are complex and deep learning results poorly due to insufficient data.

Images