Zijian Dong
@dong_zijian
PhD student at ETH Zurich, focus on 3D Computer Vision and Virtual Human.
**Introducing AG3D**: Learning to Generate 3D Avatars from 2D Image Collections. 🔗Full Info: zj-dong.github.io/AG3D/ Catch our presentation at @ICCVConference in Paris! 🗓️Date: Thursday 🕰️Time: 14:30 - 16:30 📌Paper ID: 1836 📍Location: Room "Foyer Sud" 088 🔽Details Below:
[Poster 277 Thu-PM] AvatarPose estimates 3D poses and shapes of multiple closely interacting people from multi-view videos. We achive this by learning personalized avatars and refine poses with the avatars. Details: x.com/dong_zijian/st… Project: eth-ait.github.io/AvatarPose/
I am excited to share our recent research at #ECCV2024: AvatarPose: Avatar-guided 3D Pose Estimation of Close Human Interaction from Sparse Multi-view Videos. We'll present it on Thursday afternoon (session 6, id: 277). Project page and code: eth-ait.github.io/AvatarPose/
I'm happy to announce Gaussian Haircut✂️, our recent hair reconstruction work that will appear at #ECCV2024! We'll present it in the Thursday afternoon session. Project page: eth-ait.github.io/GaussianHaircut The code is coming soon!
If you are interested digital human hair/pose/clothing/interaction, you have to see this! #ECCV2024 @eccvconf 😎
Are you interested in hair reconstruction, human pose and shape estimation, digital human with clothing deformation, and hand-object interaction? Checkout our ECCV papers in the poster session at #ECCV2024! See 🧵.
The personalized avatar prior also enables us to reduce the collision between interacting people:
Struggling to secure more GPUs for training large X (reconstruction, Gaussian, etc) models? Check out LaRa, a lightweight 3D vision model designed to efficiently handle large-baseline reconstruction challenges apchenstu.github.io/LaRa/
Are you interested in hair reconstruction, human pose and shape estimation, digital human with clothing deformation, and hand-object interaction? Checkout our ECCV papers in the poster session at #ECCV2024! See 🧵.
🤔What about using an LLM as a function approximator for f(x; θ) where the parameters θ are natural language? 🤔Can we learn θ just like in machine learning (ML) where θ are numerical values? ✨Check out Verbalized ML, where data and models both operate in natural language! 🤩
Our @CVPR #CVPR24 paper, TransFusion, integrates a textual summary of the past with a single frame to predict next object interactions in egocentric videos and achieves sota performance on the #Ego4D benchmark. Project page: eth-ait.github.io/transfusion-pr…
🚀 #CVPR2024 is almost here! Next week, AIT will present 8 papers at the conference. We're excited to share overviews of each project right here. Looking forward to connecting with you all in Seattle! 🌟 👇
Need more than static 3D models? At #CVPR2024, we present Dream-in-4D, a unified approach for text-to-4D, image-to-4D and personalized 4D generation. With video diffusion prior, we learn plausible motions for 3D assets. Work partially done during an internship at NVIDIA.