Texture is among the most important visual properties that determine the appearance of materials, and therefore plays a pivotal role in this context. As a consequence, texture analysis has been an area of intense research activity in computer vision since early on.
In recent years, research in the field has been shifting from 'hand-designed' descriptors towards data-driven, 'learnt' approaches (Deep Learning). The applicability of this paradigm to textures analysis, recognition and retrieval, however, is yet to be understood. It is also unclear how previous knowledge from the 'hand-crafted' era can be conveniently integrated with Deep Learning models to produce flexible, robust and computationally cheap image descriptors for texture recognition.
The aim of this workshop is to provide a forum for scientists and practitioners concerned with theory and applications of texture analysis, classification and retrieval. We welcome original contributions that address theoretical and practical issues including, but not limited to:
- Feature design: Hand-crafted methods, deep learning and hybrid approaches;
- Vision Science: Perception models for textures;
- Mathematical foundations of texture analysis;
- Performance assessment: Databases under controlled conditions, classification in the wild, comparative evaluations;
- Applications: Material recognition, Remote sensing, medical image analysis, food recognition, fashion industry and industrial inspection, cultural heritage.