Published Paper Details:

The AI-powered virtual try-on system represents a transformative approach in the fashion industry by enabling consumers to visualize clothing items on themselves digitally. This system utilizes state-of-the-art Generative Adversarial Networks (GANs) for realistic image synthesis and Long Short-Term Memory (LSTM) networks to generate personalized fashion recommendation, the system addresses key challenges in e-commerce such as fitting uncertainty and personalization. This system improves user experience by generating high-fidelity, pose-adaptive try-on images and recommending garments based on temporal user behavior patterns. Implemented via a modular GAN-LSTM pipeline, this technology aims to minimize the gap between physical and online shopping by allowing consumers to virtually try garments tailored to their body shapes and preferences. The system's dual focus on visual realism and behavioral personalization improves user engagement, reduces return rates, and facilitates data-driven fashion insights. Experimental results demonstrate notable improvements in image realism and recommendation accuracy.