Attention‐based vector quantisation variational autoencoder for colour‐patterned fabrics defect detection

Abstract

Defect detection is an essential link in the fabric production process. Due to the diversity of patterns and scarcity of defect samples for colour‐patterned fabrics, reconstruction‐based unsupervised deep learning algorithms have received extensive attention in the field of fabric defect detection. Among them, unsupervised reconstruction models based on variational autoencoders (VAEs) have been shown to be effective. However, there is a problem of posterior collapse in the process of modelling parametric distributions of continuous variables by VAEs. Therefore, VAE‐based defect detection methods for colour‐patterned fabrics usually produce ambiguous reconstruction results, thereby affecting the defect detection performance. In this article, an attention‐based vector quantisation variational autoencoder (AVQ‐VAE) is proposed for colour‐patterned fabric defect detection. The method adopts autoregressive modelling of discrete variables to avoid the posterior collapse problem of traditional VAEs, and utilises attention mechanism to enhance the feature representation ability of the model. AVQ‐VAE consists of encoder, embedding space, decoder and attention mechanism. The encoder is used to map the input image into multiple feature vectors. Vector quantisation in embedding space is used for discretisation and autoregressive modelling of feature vectors. A decoder is used to decode discrete variables into images of the same size as the original input. Furthermore, an attention mechanism is used to capture channel and spatial correlations, which help the model focus on important information by adaptively recalibrating feature maps. Experimental results on public datasets demonstrate that the proposed method is robust and effective for colour‐patterned fabric defect detection.

Evolving regularised random vector functional link by seagull optimisation algorithm for yarn‐dyed fabric colour difference classification

“…Figure 9 shows the confusion matrix of the MSOA‐RRVFL model in the first classification, and the classification parameters are set as follows: number of hidden ganglia: 90; activation function: Sigmod; population number: 20; maximum iterations: 60; training sample size: 1040; and test sample size: 520. According to the colour difference classification standard, 4 this paper divides the test samples into five categories: microscopic, trivial, noticeable, very noticeable and intense colour difference. The first category: 47/520; the second category: 168/520; the third category: 150/520; the fourth category: 119/520; and the fifth category: 36/520.…”

“…Fabric defect detection is also a crucial stage in the manufacturing of textiles. Zhang et al 4 proposed an attention‐based vector quantisation variational automatic encoder model for the detection of colour pattern fabric defects. In addition, Zhang et al 5 proposed an approach for fabric defect prediction using unsupervised and commemorative defect features, considering the weak generalisation ability of autoencoders.…”

A Timestep-Adaptive-Diffusion-Model-Oriented Unsupervised Detection Method for Fabric Surface Defects

“…In terms of image generation, GAN outperforms the AE-based method [13], increasingly extending its application in derived models for defect detection. Zhang et al [24] integrated attention mechanisms based on GAN to enhance its feature representation capability for high-quality information, achieving better reconstruction. Wei et al [25] conducted multistage training based on a deep convolutional generation adversarial network (DCGAN) and reduced the interference of defects in the image reconstruction using the linear weighted integration method.…”

Progressive mask‐oriented unsupervised fabric defect detection under background repair

“…Since obtaining non‐defect samples during production is quite simple, unsupervised anomaly detection methods are gaining popularity. Zhang et al 14 introduced the attention mechanism based on Variational Auto‐Encoder (VAE) for enhancing defect detection performance, adopted the autoregressive model of the discrete variables to make the reconstruction result clearer, and effectively avoided the posterior collapse problem in VAE. Hu et al 15 developed a convolutional denoising auto‐encoder and hash encoder to extract the features of fabric prints and minimise the detection time.…”