In the textile production chain, fibers serve as the foundational units for yarn, and yarn, in turn, acts as a fundamental component for woven or knitted fabrics. The quality control of fabrics is intricately tied to the management of fibers and yarns. Traditional laboratory methods have been utilized to assess their quality, but the advent of machine learning and deep learning introduces a transformative approach. This review explores the application of machine learning methods such as principal component analysis, support vector machine, and deep learning methods such as artificial neural networks, convolutional neural networks, you look only once, and genetic algorithms to predict various properties of fibers and yarns. In the context of fibers, the review delves into topics such as cotton fiber grading based on color, characterization of jute fiber, and the identification of medullation in alpaca fibers. For yarns, the focus shifts to predicting parameters such as yarn tenacity, evenness, abrasion index of spun yarns, inspection of false twist textured yarn packages, breaking elongation of ring-spun cotton yarns, tensile properties of cotton/spandex yarns, yarn thickness, and yarn hairiness. The review also provides insights into the advantages and limitations of the discussed studies. Despite the comprehensiveness of this review, it is acknowledged that there might be additional relevant work not covered. The review encourages the sharing of data to expedite the integration of these technologies in future applications within the field.
