What’s your problem statement?

Fashion consumers struggle to find outfits that reflect their personal styles and cultural identities due to static e-commerce platforms and overwhelming choices. Traditional approaches lack personalization and responsiveness, leading to frustration. FashionGen aims to solve this by leveraging Generative AI for personalized outfit recommendations in a conversational format, enhancing the online shopping experience while considering individual preferences and current trends.

The problem FashionGen solves

FashionGen addresses key challenges in online fashion shopping by providing personalized outfit recommendations tailored to individual preferences and cultural identities. Users can engage with a Conversational Fashion Outfit Generator, offering an interactive experience with real-time feedback and adjustments. The platform features a Virtual Try-On option, allowing users to visualize outfits before purchasing, enhancing their confidence in choices. Additionally, FashionGen continuously adapts to current trends by learning from social media and festive fashion highlights. With seamless integration of Amazon product links, users can easily access and purchase recommended items, creating a streamlined, personalized, and culturally relevant shopping experience.

Challenges we ran into

Integrating multiple models into FashionGen presented several challenges that required strategic solutions. Ensuring compatibility between Generative AI for outfit recommendations and Stable Diffusion for virtual try-ons was crucial, necessitating the standardization of data formats and communication protocols. Achieving real-time responsiveness for the conversational interface demanded optimization to minimize latency, allowing users to receive prompt recommendations and feedback. Additionally, managing and synchronizing data across various models posed challenges due to differing data requirements; we developed a robust data pipeline to ensure all models accessed relevant, updated information.