Pedestrian attribute recognition method, device, electronic equipment and storage medium

Abstract

Embodiments of the present invention provide a pedestrian attribute recognition method, device, electronic equipment, and storage medium. The method includes: obtaining a multi-category convolutional neural network model; using convolutional neural network technology, the multi-category convolutional neural network model is The number of pedestrian attribute categories is reconstructed to obtain the first pedestrian attribute recognition model. The reconstruction includes the reconstruction of the storage data format; through transfer learning, the first pedestrian attribute recognition model is trained using the pre-prepared pedestrian attribute data set. obtaining a second pedestrian attribute recognition model; optimizing the second pedestrian attribute recognition model through a preset programmable reasoning accelerator to obtain a target pedestrian attribute recognition model; performing pedestrian attribute recognition on an image to be recognized through the target pedestrian attribute recognition model to obtain Pedestrian attribute recognition results. The invention solves the problem of increasing the running speed of the model under the condition that the accuracy of the model loss is small by optimizing and accelerating the pedestrian attribute recognition model.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a pedestrian attribute recognition method, device, electronic equipment and storage medium. Background technique [0002] In recent years, pedestrian attribute recognition has been more and more widely used in various fields. Accurate and efficient pedestrian attribute recognition results play an important role in assisting managers in handling incidents and improving human-computer interaction experience. Especially with the development of deep learning technology in the field of artificial intelligence and the improvement of the computing power of related hardware equipment, building a deep convolutional neural network model to extract features to solve the problem of pedestrian attribute recognition has gradually become the main technical means, which is of great importance to improve pedestrians. The accuracy of attribute recognition plays a positive role. But the more...

AI Technical Summary

Problems solved by technology

[0003] The embodiment of the present invention provides a pedestrian attribute recognition method, which can solve the problem of increasing the running speed of the model when the accuracy of the model loss is small by optimizing and accelerating the pedestrian attribute recognition model, which is beneficial to large-scale pedestrians in the production environment. Improve the processing efficiency of pedestrian attribute recognition when using large-scale data, which can greatly help save time and cost and improve related economic and social benefits

Images