Deep neural network training method and device, electronic equipment and storage medium

Abstract

The invention discloses a deep neural network training method, belongs to the technical field of computers, and is used for solving the problem of relatively low performance of a trained neural network in a complex scene in the prior art. The method comprises the following steps: obtaining a plurality of training samples provided with preset category labels, and training a neural network model based on the plurality of training samples; wherein the loss function of the neural network model is used for carrying out weighting operation according to a first weight value in direct proportion to the distinguishing difficulty of each training sample, and determining the loss value of the neural network model. According to the deep neural network training method disclosed by the embodiment of theinvention, the importance of the training samples with higher difficulty in distinguishing in the training samples is adaptively improved, so that the samples with higher difficulty in distinguishingare prevented from being mistakenly classified by the neural network obtained by training, and the performance of the neural network is improved.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a deep neural network training method, device, electronic equipment and storage medium. Background technique [0002] In recent years, deep learning has made remarkable progress in the field of pattern recognition, and its key factors may include: rich and flexible network models, strong computing power, and better adaptability to big data processing. As neural networks are applied to different tasks, the improvement of neural network models is also a key issue for those skilled in the art to study. The improvement of the neural network model in the prior art mainly focuses on two aspects: network structure and loss function. Among them, the loss function commonly used in classification model training is mainly Softmax loss, and Center loss based on Softmax improvement at the top international conference ECCV in 2016. However, the applicant found through research on...

AI Technical Summary

Problems solved by technology

However, the applicant found through research on the neural network using Center loss as the loss function in the prior art that if there are samples with larger noise or weaker discrimination in the training samples of the training neural network, using the existing The model trained by the loss function is tested, and the improvement of the classification or recognition results of the trained neural network is limited.

Images