[CVPR'22] ICON: Implicit Clothed humans Obtained from Normals

[CVPR'23, Highlight] ECON: Explicit Clothed humans Optimized via Normal integration

ICASSP 2022: "Text2Video: text-driven talking-head video synthesis with phonetic dictionary".

AI based all-in-one character generator Blender plug-in. This project contains unofficial updates for the CEB_ECON Blender add-on.

Python implementation accompanying the Transformer Inertial Poser paper at SIGGRAPH Asia 2022

Towards Robust 3D Body Mesh Inference of Partially-observed Humans: semester project at ETH Zurich

Emotionally responsive Virtual Metahuman CV with Real-Time User Facial Emotion Detection (Unreal Engine 5).

[CVPR 2025 Highlight] Real-time Free-view Human Rendering from Sparse-view RGB Videos using Double Unprojected Textures

[ECCV 2024] MetaCap: Meta-learning Priors from Multi-View Imagery for Sparse-view Human Performance Capture and Rendering

Official PyTorch Implementation of "Generative Approach for Probabilistic Human Mesh Recovery using Diffusion Models", ICCV 2023 CV4Metaverse Workshop

Official repository of "Camera Distortion-aware 3D Human Pose Estimation in Video with Optimization-based Meta-Learning", ICCV 2021

PyTorch implementation of the human neural rendering in unseen positions presented at WACV 2022 "Creating and Reenacting Controllable 3D Humans with Differentiable Rendering"

[WACV 2026] Official code repo of paper ArchitectHead: Continuous Level of Detail Control for 3D Gaussian Head Avatars.

The Virtual Human Toolkit enables researchers and developers to rapidly create embodied conversational agents using speech recognition, audio-visual sensing, natural language processing, text-to-speech synthesis, and nonverbal behavior generation.

The AVATAR project demonstrates virtual characters that play the role of a police officer or a witness in a police interview.

Prototype of a narrative-memory virtual human: identity emerges from stories, not sliders. Durable anchors (event facts) plus adaptive retellings (“flux”) create a coherent, context-sensitive personality.

Simple class to work with meshes

🎥 Create high-quality explainer videos from keywords or Markdown with Text2Video's AI-driven automation and flexible deployment options.

3D body capture with enriched automatic modelling. (WORK IN PROGRESS)