Pages 72–78
Abstract
Research summary supplied by the authors
Colours serve as an embodiment of preference, choice, identity, expression and aesthetics. The colours of the products, we carry have the power to represent ourselves. The strategic choice of colours in product manufacturing plays a pivotal role in elevating the appeal, identity, style and value of the product. In the Leather context, the end products are highly dependent on the fashion colour trends as they need validation from the fashion industry. Seasonal colour forecasting namely Fall Winter (FW) and Spring Summer (SS) play a huge role in leather industry as it acts as the guidelines for the manufacturers that helps them to align their productions with global consumer preferences and demands. It also indirectly aids in streamlining the supply chain management by contributing to the enhanced logistics. The primary objective of the work is to build a model that predicts the seasonal colour trends by taking multiple data sources having historical colours. The approach involves optimization algorithm namely Hungarian matching to segregate the input based on minimum Delta E difference and feeding the segregated colour inputs to Deep Learning model that forecasts the upcoming colours of the season. The outputs from different data source models are further fed to an ensemble ML technique namely Consensus-Dispersion Weighted Ensemble that fuses the output from different data source models to a unified output that has improved accuracy. The findings demonstrate that the AI-driven models can effectively learn slow seasonal colour transitions, and can ease the early prediction of emerging hues which can benefit the manufacturers and designers. It overcomes the manual colour prediction systems as they can be biased due to human perceptions and subjectivity. This work significantly contributes to the leather sector by enabling to make informed decisions early in the product development. The system supports better material planning, reduces wastage, and strengthens competitiveness by offering scientifically grounded future colour predictions catering to evolving consumer preferences and global fashion dynamics.