Clothing material classification-based deep learning

Aiming at the problems of poor classification and recognition of traditional clothing images and low recognition accuracy, this paper proposes a convolutional neural network clothing material classification and recognition method based on the improved deep residual network (ResNet). By introducing an attention mechanism module in each residual block, a clothing material image classification and recognition model based on improved ResNet is constructed. In order to better carry out the related research, a self-constructed dataset specifically used for clothing material classification and recognition. In order to verify the superiority of the improved network in clothing material recognition, the traditional machine learning methods and the classical neural networks in deep learning are explored respectively, and the parameters of the network are adjusted to make it more suitable for clothing material recognition. The improved model is compared with other models. The experimental results show that the performance of the clothing material recognition network based on the improved ResNet network is superior (99.2% recognition accuracy, 98.46%precision, 98.83%recall, and 0.9864 F1_Score), and it can satisfy the needs of commercial networks for recognizing clothing materials