A repository to curate and summarise research papers related to fashion and e-commerce

👔👗🕶️🎩 Cool resources about Fashion + AI! (papers, datasets, workshops, companies, ...) (constantly updating)

Open-source APIs, SDKs, and models for building virtual try-on and fashion AI applications. Generate models, edit garments, create photoshoots, and build personalized fashion experiences.

Full pipeline for TianChi FashionAI clothes keypoints detection compitetion in TensorFlow

dresses segmentation from any background

[ACM TIST 2025] GenAI in Fashion: Overview, also includes 🔥latest papers, ⚙️metrics, 👀workshops, 🚀companies & products, ...)

FashionAI Key Points Detection using CPN model in Pytorch

No description available

Code for TianChi 2018 FashionAI Cloth KeyPoint Detection Challenge

Fashion Matrix is dedicated to bridging various visual and language models and continuously refining its capabilities as a comprehensive fashion AI assistant. This project will continue to update new features and optimization effects.

FashionAI Global Challenge—Attributes Recognition of Apparel—Ranked 21st solution.

Mask R-CNN for FashionAi key point location

fAshIon Datasets

FashionAI Clothes Attribute Recognition

The Fashion AI Assistant is a recommendation system that combines vector search capabilities with multimodal AI to provide intelligent fashion recommendations.

FashionAI Global Challenge—Attributes Recognition of Apparel based on PyTorch

Extract attribute of clothes detected in image.

AI agents and Nix: parametrable skills/instructions and tools, packaged together in a reproducible and modular fashion

AI fashion design pipeline turning text prompts into runway-ready sketches and videos. 🎨

Heatmap approach for Fashion AI keypoint Challenge

FashionGPT is an AI-powered product recommender system designed to help users find their perfect fashion match. By leveraging the power of Transfer Learning, FashionGPT can analyze the visual features of a given image and recommend similar products that cater to the user's unique style preferences.

No description available

No description available

No description available

A Fashion AI app that styles and suggest outfits to wear with clothes in your actual wardrobe

StylistAI revolutionizes fashion with AI-powered solutions. Featuring a Text-To-Outfit Generator, Human Identification, Outfit Recommendation, Virtual Try-On, and a Fashion Chatbot, this platform offers personalized outfit suggestions, virtual fitting experiences, and interactive fashion advice.

👕👖🥻Implementations of AI for my UG Project-"Generative Models and Recommender Systems for AI-driven Fashion"

RL STaR is a platform for creating AI for robotic applications. Researchers can share their work in the form of new robots or new tasks. Everything was made in a modular fashion to be easy to use and easy to share.

AI-powered fashion recommendation system leveraging LLMs, embeddings, and retrieval techniques to deliver personalized shopping experiences.

Artistic style transfer has been part of the quickly growing AI Art community in recent times. Pioneered by Gatys et al this class of methods allows for the transfer of a style, texture, pattern ect. to a target image. This has been made possible by the use of the expressive hidden features of large computer vision models like VGG19 to use as a loss function. The method by which this transformation is undertaken has come in many forms from the intial method which optimized the pixel of the image directly to newer forms that train some form of CNN to create a general style transfer network. The method presented in this paper is in the same vain as those newer methods but leveraging a new class of deep learning method, that of deep neural cellular automata. This new method brings with it the ability to transform any image to a target style, like the CNN method mentioned earlier, using the same automata update rule in an iterative fashion. This iterative fashion allows one to stop the process early for less stylization as well as resulting in video rather than a still image which, due to the dynamic nature of cellular automata, has visually pleasant properties.