Revolutionizing Handloom Design Creation Through Associated Feature Transformation and Advanced Deep Decoder Techniques
Anindita Das, Aniruddha Deka · 2023
With a particular focus on handloom attire, this research explores a deep-learning technique for creating designs. Due to the difficulty to generalize the style and content mapping that impacted visual quality, existing deep neural-based approaches are constrained. In this work, a multimodal network is employed incorporating the multi-layered neural network with feature transferring and a deep decoder network. The technique illustrates how adequate quantitative approaches can be used to produce high-quality handloom designs. Additionally, this approach offers a fresh dataset for the next automatic handloom systems.