Brain-inspired long-time-interval continuous pedestrian re-identification method and device

Abstract

The invention discloses a brain-inspired long-term continuous pedestrian re-recognition method and device, and the method comprises the steps: training a deep neural network through employing initial training data, and obtaining an initial pedestrian re-recognition model; extracting scene memory buffer data in the initial training data, and obtaining preset parameters in the initial pedestrian re-recognition model; when the initial pedestrian re-identification model is retrained according to the newly added training data and the scene memory buffer data, preset parameters are constrained, so that the preset parameters meet constraint conditions; and calculating a loss function of the retrained initial pedestrian re-identification model, and performing iterative updating on the scene memory buffer data and preset parameters according to the loss function until the current pedestrian re-identification model meets an iteration termination condition, thereby obtaining a final pedestrian re-identification model. According to the method, the neural network model parameters can be updated under the condition that a large amount of old data does not need to be stored, and the performance of the model on the old data is maintained while new data is adapted.

Description

technical field [0001] This application relates to the technical field of pedestrian re-identification based on deep learning, in particular to a brain-inspired long-term continuous pedestrian re-identification method and device. Background technique [0002] Pedestrian re-identification technology is a computer vision technology whose purpose is to retrieve pedestrians with specified identities in the data captured by different cameras, that is, according to the image of the specific identity pedestrian being queried, from the candidate pedestrian gallery captured by multiple different cameras Choose images that are similar to its characteristics. Pedestrian re-identification technology plays a vital role in the fields of intelligent monitoring and security. [0003] Nowadays, deep learning methods have been widely used in the field of pedestrian re-identification. The general process is to input the pedestrian image to be queried into the deep neural network model, extrac...

AI Technical Summary

Problems solved by technology

[0006] This application provides a brain-inspired long-term continuous pedestrian re-identification method and device to solve the problem of catastrophic forgetting in the related technology-based deep learning-based pedestrian re-identification method when continuously learning new data. Update the parameters of the neural network model under the condition of storing a large amount of old data, and maintain the performance of the model on the old data while adapting to the new data

Images