Person-vehicle weight recognition and its model training method, device, equipment and storage medium

Abstract

The present application discloses a person-vehicle weight recognition and its model training method, device, equipment and storage medium, wherein, the person-vehicle weight recognition model training method includes: obtaining a plurality of training samples, each training sample including the first This image and the second sample image; use the training samples to train the pre-established initial person-vehicle weight recognition model until the loss value of the loss function in the initial person-vehicle weight recognition model meets the preset conditions, and obtain the target person-vehicle weight recognition model ; Wherein, the initial human-vehicle recognition model includes a target detection model, a first initial multi-layer perceptron and a second initial multi-layer perceptron. In the embodiment of the present application, when training the initial human-vehicle weight recognition model, at least two sample images can be used to enhance the features of each other, and the feature information of the first object in different sample images can be fully utilized to improve the image quality caused by the change of shooting state. The loss of feature information helps to improve the re-identification effect of people and vehicles.

Description

technical field [0001] The present application belongs to the field of information technology, and in particular relates to a method, device, equipment and storage medium for human-vehicle recognition and model training thereof. Background technique [0002] With the development of artificial intelligence technology, machine vision recognition is increasingly used in daily life, for example, pedestrian recognition based on video images collected by imaging equipment. In order to obtain pedestrian activity trajectories, pedestrian re-identification technology is usually used, that is, pedestrian recognition may be performed based on images collected by different imaging devices. [0003] Since the shooting angles of different imaging devices may be different, the state (such as length, width, or angle) of the same pedestrian in different images may be different. When the pedestrian is riding a bicycle, the combination of pedestrian and vehicle (hereinafter referred to as the ...

AI Technical Summary

Problems solved by technology

[0004] The embodiment of the present application provides a method, device, equipment and storage medium for the identification of the weight of people and vehicles and its model training, which solves the problem of poor recognition effect in the prior art when identifying the weight of people and vehicles

Images