The "Conversational Fashion Outfit Generator" is a state-of-the-art project that redefines the way users discover and curate personalized fashion outfits. In collaboration with Flipkart, we have created an innovative solution that combines the power of Generative AI and Stable Diffusion techniques to provide a seamless and engaging fashion shopping experience.
Our solution introduces a natural and interactive conversation-based interface that enables users to engage with the system using plain language. We've integrated OpenAI's GPT-4, a state-of-the-art Large Language Model, to facilitate intelligent conversations. Users can effortlessly communicate their preferences, questions, and feedback, mimicking a human-like interaction.
Our solution introduces a natural and interactive conversation-based interface that enables users to engage with the system using plain language. We've integrated OpenAI's GPT-4, a cutting-edge Large Language Model, to facilitate intelligent conversations. Users can effortlessly communicate their preferences, ask questions, and provide feedback, mimicking a human-like interaction.
Leveraging a specially curated dataset containing user details from Flipkart, we employ Universal Named Entity Recognition (UniversalNER) to extract valuable insights. This allows us to understand user-specific style choices, color preferences, favorite brands, and shopping history. These preferences serve as the foundation for our outfit recommendations.
Our Fashion AI assistant generates personalized outfit recommendations based on the user's preferences and interactions. By integrating Stable Diffusion for Image Manipulation, users can upload their photos to virtually "try on" different outfits. This interactive experience bridges the gap between visualization and reality, enhancing the user's confidence in their fashion choices.
We ensure that our recommendations stay in sync with the latest fashion trends by regularly updating the system's knowledge base. Our feedback loop collects data from reputable fashion news sources, blogs, social media influencers, and review websites. This real-time trend analysis, combined with user preferences, results in outfit suggestions that are both personalized and in vogue.
To ensure the coherence and compatibility of the suggested outfits, our solution considers a myriad of factors. These include the user's body type, the occasion for which the outfit is intended, regional preferences, and age group. Advanced diffusion models assist in intelligently customizing the recommendations for each user's unique needs.
We empower users to actively participate in the outfit recommendation process. The conversational interface allows users to provide feedback on recommended outfits. Additionally, users can modify and fine-tune outfits, fostering a dynamic and collaborative shopping experience.
Our project presents a complete Flipkart clone website that replicates the actual shopping experience. A standout feature is the "Try it on" AI capability, enabling users to virtually wear outfits before purchasing. This innovation enhances user engagement, satisfaction, and the likelihood of making informed buying decisions.
Our solution employs a sophisticated blend of AI models. This includes Control-net for preserving the pose of uploaded images, SegFormer for accurate clothes segmentation, and LLMs fine-tuned to serve as the Fashion AI assistant. The AI ecosystem ensures accurate understanding, image manipulation, and tailored recommendations.
To provide users with a clearer view of recommended outfits, we utilize the "doevent/Face-Real-ESRGAN" model for image upscaling. This enhancement contributes to a high-quality and immersive try-on experience.
To run this project, you can use Google Colab. We've prepared a Google Colab notebook with all the necessary code and instructions. Simply click the link below to open the notebook:
The Conversational Fashion Outfit Generator powered by GenAI revolutionizes the way users discover and engage with fashion products. It bridges the gap between technology and fashion, providing a seamless blend of AI-driven personalization, real-time trends, and interactive experiences. Users are empowered to make confident fashion choices, explore their preferences, and stay up-to-date with the latest fashion trends. The project not only enhances the shopping experience but also represents a remarkable convergence of AI capabilities and the fashion industry.
The Conversational Fashion Outfit Generator powered by GenAI revolutionizes the way users discover and engage with fashion products. It bridges the gap between technology and fashion, providing a seamless blend of AI-driven personalization, real-time trends, and interactive experiences. Users are empowered to make confident fashion choices, explore their preferences, and stay up-to-date with the latest fashion trends. The project not only enhances the shopping experience but also represents a remarkable convergence of AI capabilities and the fashion industry.
More Results at: Results
We allow try on the go feature which allows user to change the designs of there clothes smoothly and effortlessly.
