camenduru
@camenduru
building 🥪 @tost_ai ❤ open source http://github.com/camenduru
I’m back from an amazing, long long multi-country vacation, where I learned a lot about Life, the Universe, and Everything. 🌌 I’m ready to continue my open-source journey where I left off. I’m also starting my job search as an AI engineer. If you know of any opportunities or…
👀 DAViD: Data-efficient and Accurate Vision Models from Synthetic Data 🥽 Jupyter Notebook 🥳 Thanks to @fatemeh_saleh ❤ @aa_sadegh ❤ Charlie Hewitt ❤ Lohit Petikam ❤ Xiao-Xian ❤ Antonio Criminisi ❤ Thomas J. Cashman ❤ Tadas Baltrušaitis ❤ 🌐page:…
🌙 Want more curves? 😋 Flux.1 Kontext 🥪 Tost Curve Creator is now on @tost_ai 🥳 Thanks to Black Forest Labs ❤ and civitai/ultraautism ❤ 🌐page: bfl.ai/announcements/… 📄paper: arxiv.org/abs/2506.15742 📦model: civitai.com/models/1802814… 🔞 🥪tost: tost.ai
🔍 SeedVR2: One-Step Video Restoration via Diffusion Adversarial Post-Training 🎥 Jupyter Notebook 🥳 Thanks to @Iceclearwjy ❤ Shanchuan Lin ❤ Zhijie Lin ❤ Yuxi Ren ❤ Meng Wei ❤ Zongsheng Yue ❤ Shangchen Zhou ❤ Hao Chen ❤ Yang Zhao ❤ Ceyuan Yang ❤ Xuefeng Xiao ❤ Chen…
👁️ ObjectClear: Complete Object Removal via Object-Effect Attention 🧹 Jupyter Notebook 🥳 Thanks to Jixin Zhao ❤ @ShangchenZhou ❤ @ZhouXia1212 ❤ @peiqing001 ❤ @ccloy ❤ 🌐page: zjx0101.github.io/projects/Objec… 🧬code: github.com/zjx0101/Object… 📄paper: arxiv.org/abs/2505.22636…
🎞️ Wan 2.1 T2V 🎥 Classic 90s Film Aesthetic is now on @tost_ai 🥳 Thanks to Team Wan ❤ 🌐page: wan.video/welcome 🧬code: github.com/Wan-Video/Wan2… 📄paper: arxiv.org/abs/2503.20314 🍇runpod: github.com/camenduru/wan2… 🥪tost: please try it 🐣 tost.ai
🧵 Wan2.1 Text To Video - Classic 90s Film Aesthetic - The Crow Style 🤍 📦model: civitai.com/models/1773251…
🎙️ FLOAT: Generative Motion Latent Flow Matching for Audio-driven Talking Portrait 👄 Jupyter Notebook 🥳 Thanks to @taekyungki ❤ @DongchanM ❤ Gyeongsu Chae ❤ Thanks to @bk_sakurai ❤ for the tutorial. 🌐page: deepbrainai-research.github.io/float/ 🧬code: github.com/deepbrainai-re… 📄paper:…
(1/n) 🎉Excited to share that FLOAT has been accepted to #ICCV2025! It can generate high-fidelity talking portrait videos in real-time and supports test-time head pose editing via the learned motion orthonormal basis. We’ve released the code and weights — enjoy exploring FLOAT!
🧩 PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers 📐 Jupyter Notebook 🥳 Thanks to @kevin_yuchenlin ❤ @lin_chenguo ❤ @paulpanwang ❤ Honglei Yan ❤ Yiqiang Feng ❤ Yadong Mu ❤ Katerina Fragkiadaki ❤ Thanks to @alexandernasa ❤…
🚀 Meet #RadialAttention — a static sparse attention mechanism with O(nlogn) complexity for long video generation! ✅ Plug-and-play: works with pretrained models like #Wan, #HunyuanVideo, #Mochi ✅ Speeds up both training&inference by 2–4×, without quality loss 🧵1/4
🚀 Our #ICCV25 paper "GO TO ZERO" is out - the first systematic exploration of zero-shot text-to-motion generation! ✅ 2M+ high-quality 3D motions ✅ New benchmark with 126 real-world prompts ✅ 7B motion foundation model We're pushing motion generation into the zero-shot era!
🚨 We just released 🎞️MoVieS — a feed-forward model that reconstructs 4D scenes in ⚡️1 second My favorite part: It learns dense (pixel-wise) sharp 3D world movements from novel view rendering + sparse point tracking supervision 🤯🎯 Check it out 👉 chenguolin.github.io/projects/MoVieS