Batik Recognition and Classification Using Transfer Learning and MobileNet Approach

In the vibrant tapestry of Indonesian culture, Batik motifs stand out as a testament to its enduring artistic heritage. Yet, adapting these intricate patterns, particularly the mesmerizing "insang" featuring fish-like forms, presents a unique challenge for modern applications. Limited datasets and the need for efficient mobile solutions create a bottleneck in accurate motif classification. This research boldly tackles this challenge by proposing a groundbreaking approach: marrying the power of MobileNet architecture, specifically designed for mobile devices, with transfer learning techniques. Transfer learning acts as a bridge, leveraging knowledge from a vast dataset to compensate for limited data specific to Batik. This synergy unlocks remarkable accuracy, with our method achieving a stunning 98% classification rate in under a second on mobile devices. The implications of this breakthrough are far-reaching. It safeguards Batik's legacy by enabling its digital preservation and paves the way for its seamless integration into contemporary design. It is predicted that Batik motifs can adorn digital interfaces, enrich user experiences, and inspire innovative fashion trends. This research is a beacon illuminating the path for Batik to evolve and thrive in the digital age. By empowering mobile devices to recognize and interpret these intricate patterns, it aims to unlock many possibilities. Batik's rich history can be woven into the fabric of modern life, enriching our digital landscapes and fostering a deeper appreciation for this cultural gem. This is not merely a technological feat; it is a celebration of tradition, a bridge between generations, and a testament to the enduring power of creativity.