#ECCV2024 Presenting a poster was truly a joyful experience, and listening to other posters broadened my horizons! I met many old friends and made many new ones. Looking forward to future reunions!

Introducing MarDini, a model built from scratch to combine the strengths of diffusion and masked auto-regressive approaches for large-scale video generation.

Introducing MarDini 🍸 -- our latest exploration in video diffusion models from AI at Meta! MarDini brings an asymmetric design that breaks down video modelling into two sub-tasks: 1. A masked, auto-regressive, heavy-weight planning model focusing on long-range temporal

🚀Introducing Leffa: Learning Flow Fields in Attention for Controllable Person Image Generation! 👤🔄👗 Leffa is a unified framework for controllable person image generation that enables precise manipulation of both appearance (virtual try-on) and pose (pose transfer). It guides

Meta releases Leffa Learning Flow Fields in Attention for Controllable Person Image Generation

【MetaがLeffaをリリース！？人物画像生成の新時代！】 ✎. FYIG: Metaが、制御可能な人物画像生成技術「Leffa」をリリースしたそうです！ Leffaは、"Learning Flow Fields in Attention for Controllable Person Image Generation"の略で、

Leffa from Meta is a 🆕 unified framework for controllable person image generation that enables precise manipulation of both appearance (i.e., Virtual try-on) and pose (i.e., Pose transfer). Learn more about 📚 Code,🔥 Demo, and 🤗 Model ⬇️

Thanks for sharing

Meta Launches Open-Source Virtual Try-On! Leffa is a unified framework introduced by Meta that can generate new images based on reference images, allowing users to switch between different clothing and poses. Compared to previous methods, Leffa is better at preserving

AK Amazing!!

Leffa from Meta is a 🆕 unified framework for Virtual try-on and Pose transfer!🔥 👕 > Leffa Model on Hugging Face Hub with MIT license: huggingface.co/franciszzj/Lef… > Leffa Gradio app is up on Hugging Face Spaces: huggingface.co/spaces/francis…

Meta最新开源了一款AI换装和姿势迁移的模型：Leffa，它可以控制人物的着装和姿势，细节保持得很好，生成的图像也比较自然，没有衣服变形或细节模糊的问题 换服装+换姿势的组合使用，在虚拟试衣这个层面上更强大更灵活了 #AI试衣 #虚拟换装 #Leffa #MetaLeffa

Leffa：Meta开源的AI试衣模型 效果非常好，可以媲美可灵的AI换衣效果。支持AI换衣或者调整人物动作。 在线体验：huggingface.co/spaces/francis… Github：github.com/franciszzj/Lef… 模型：huggingface.co/franciszzj/Lef…

New Virtual TryOn model on @fal! Leffa is a higher resolution Virtual TryOn model that can be used commercially and provides better outputs than CatVTON! Video generated with Hailuo AI (MiniMax)'s Video-01-Live model! Workflow in 🧵

🚀🚀🚀 We have made some updates on Leffa, and it now takes only 6 seconds (on A100) to generate an image (previously it took 45 seconds). Feel free to give it a try! Demo: huggingface.co/spaces/francis… Code: github.com/franciszzj/Lef…

Success is uncertain, but entertainment is guaranteed! ✨

Excited to share that Leffa has been accepted to CVPR 2025 🚀🚀🚀 Paper: arxiv.org/abs/2412.08486 Code: github.com/franciszzj/Lef… Demo: huggingface.co/spaces/francis… Model: huggingface.co/franciszzj/Lef…

#CVPR25 Warmly welcome everyone to stop by Poster #223 in ExHall D on the morning of June 13 for a friendly chat and engaging discussion!