Pedestrian re-identification method based on natural language description

Abstract

The invention discloses a pedestrian re-identification method based on natural language description and relates to processing of a recording carrier for identifying graphics. Specifically, the invention discloses an image and a natural language are designed to describe a double-branch network structure. In the image branch network structure a MobileNet convolutional network is adopted to carry out image feature extraction to extract text features of a natural language description branch network structure through a BiLSTM network. a stacking loss function is created for similarity measurementparts between image features and text features, and network training is carried out. A corresponding pedestrian image contained in the to-be-detected image set is searched by using a trained network to achieve pedestrian re-identification based on natural language description of a stacking loss function . The defects that in the prior art, the text feature characterization of a feature extractionpart is not high, a loss function part is difficult to train a network for a long time, and a large amount of memory is consumed in the training process are overcome.

Description

technical field [0001] The technical solution of the present invention relates to the processing for identifying graphic record carriers, in particular to a pedestrian re-identification method based on natural language description. Background technique [0002] In surveillance video, due to camera resolution and shooting angle, it is usually impossible to obtain very high-quality face pictures. In the case of face recognition failure, pedestrian re-identification has become a very important alternative technology. With the help of massive image data captured by a large number of surveillance cameras, pedestrian re-identification technology can achieve pedestrian re-identification tasks relatively accurately within a certain range, and even discover and locate the current position of pedestrians in real time, which is of great significance in the field of public security. . However, in practical applications, not all tasks that require pedestrian re-identification have imag...

AI Technical Summary

Problems solved by technology

[0007] The technical problem to be solved by the present invention is to provide a pedestrian re-identification method based on natural language description, which is to design a double-branch network structure for image and natural language description. The image branch network structure uses MobileNet convolutional network for image feature extraction, natural language Describe the branch network structure to extract text features through the BiLSTM network, build a stacking loss function for the similarity measurement part between image features and text features, and perform network training. After the obtained training model is imported into the network, search all The corresponding pedestrian image included realizes the pedestrian re-identification based on the natural language description of the stacked loss function, which overcomes the problems in the existing technology that the feature extraction part of the text feature is not highly representative, the loss function part is difficult to train the network, and the training time is long. The defect that the training process consumes a lot of memory

Images