FLAG (CVPR 2022)

FLAG: Flow-based 3D Avatar Generation
from Sparse Observations

International Conference on Computer Vision and Pattern Recognition 2022

Sadegh Aliakbarian   Pashmina Cameron   Federica Bogo   Andrew Fitzgibbon   Thomas J. Cashman

arXiv (coming soon) Video (coming soon) Supplementary material (coming soon)


To represent people in mixed reality applications for collaboration and communication, we need to generate realistic and faithful avatar poses. However, the signal streams that can be applied for this task from head-mounted devices (HMDs) are typically limited to head pose and hand pose estimates. While these signals are valuable, they are an incomplete representation of the human body, making it challenging to generate a faithful full-body avatar. We address this challenge by developing a flow-based generative model of the 3D human body from sparse observations, wherein we learn not only a conditional distribution of 3D human pose, but also a probabilistic mapping from observations to the latent space from which we can generate a plausible pose along with uncertainty estimates for the joints. We show that our approach is not only a strong predictive model, but can also act as an efficient pose prior in different optimization settings where a good initial latent code plays a major role.

FLAG's Overview

FLAG with HoloLens 2


Qualitative Results (pose generation with full hand visibility)

Qualitative Results (pose generation with partial hand visibility)


    title={FLAG: Flow-based 3D Avatar Generation from Sparse Observations},
    author={Sadegh Aliakbarian and Pashmina Cameron and Federica Bogo and Andrew Fitzgibbon and Thomas J. Cashman},
    booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},