CVPR 2025

Workshop on
Photorealistic 3D Head Avatars
(P3HA)


June 11, 1pm - 6pm, 2025

Music City Center, Nashville TN

Workshop Program

Talk titles and time slots are tentative.

01:00

Opening Remarks01:00 PM - 01:15 PM

Organizers

Opening Remarks

Organizers

01:15

Invited Talk 101:15 PM - 01:50 PM

Shunsuke Saito

Codec Avatars

Shunsuke Saito
Bio

Shunsuke Saito is a Research Scientist at Meta Reality Labs Research in Pittsburgh, where he leads the effort on next generation digital humans. He obtained his PhD degree at the University of Southern California. Prior to USC, he was a Visiting Researcher at University of Pennsylvania in 2014. He obtained his BE (2013), ME (2014) in Applied Physics at Waseda University. His research lies in the intersection of computer graphics, computer vision and machine learning, especially centered around digital human, 3D reconstruction, and performance capture. His work has been published in SIGGRAPH, SIGGRAPH Asia, NeurIPS, ECCV, ICCV and CVPR, three of which have been nominated for CVPR Best Paper Award (2019, 2021) and ECCV Best Paper Award (2024). His real-time volumetric teleportation work also won Best in Show award in SIGGRAPH 2020 Real-time Live!

01:50

Invited Talk 201:50 PM - 02:25 PM

Paul Debevec

Avatars in the VFX Industry

Paul Debevec
Bio

Paul Debevec is the chief research officer at Eyeline Studios and adjunct research professor at the Viterbi School of Engineering at the University of Southern California. He obtained is PhD degree at the University of Michigan, which laid the foundation for an Acadamy Award winning VFX technique. At USC ICT Debevec has led the development of several Light Stage systems that capture and simulate how people and objects appear under real-world illumination, a technique that found countless applications in Hollywood movies. He received ACM SIGGRAPH's first Significant New Researcher Award in 2001 and in 2005 he received a Gilbreth Lectureship from the National Academy of Engineering. At Eyeline Studios Debevec continues to push the boundaries of VFX and relightable virtual avatars.

02:25

Coffee Break02:25 PM - 02:35 PM

None

Coffee Break

02:35

Invited Talk 302:35 PM - 03:10 PM

Shalini De Mello

Avatars in the Cloud

Shalini De Mello
Bio

Shalini De Mello is a Director of Research, New Experiences and a Distinguished Research Scientist at NVIDIA, where she leads the AI-Mediated Reality and Interaction Research Group. Prior to this, she was a Distinguished Research Scientist in the Learning and Perception Research Group at NVIDIA from 2013 - 2023. Her research interests are in AI, computer vision, computer graphics and digital humans. Her research focuses on using AI to re-imagine interactions between humans, and between humans and machines. She has co-authored scores of peer-reviewed publications and patents, and serves on the program committees of all major AI conferences. Her inventions have been incorporated into several NVIDIA AI products, including DriveIX, Maxine and the TAO Toolkit. She received her Doctoral and Master’s degrees in Electrical and Computer Engineering from the University of Texas at Austin.

03:10

Winner Talk 103:10 PM - 03:25 PM

None

Winner Talk 1

03:25

Winner Talk 203:25 PM - 03:40 PM

None

Winner Talk 2

03:40

Invited Talk 403:40 PM - 04:15 PM

Thabo Beeler

Life-like Avatar Animations

Thabo Beeler
Bio

Thabo Beeler is a senior staff research scientist at Google, where he is heading the Syntec team within AR Perception. He is working on digital humans in the context of virtual and augmented reality, focusing on capture, reconstruction, appearance acquisition, generative modeling, and synthesis. Previously, Thabo Beeler was a research scientist at Disney Research Studios where he built up the Capture and Effects group, focusing on digital humans for film. In 2019, his work there was rewarded with an Oscar award for technical achievements.

04:15

Coffee Break04:15 PM - 04:25 PM

None

Coffee Break

04:25

Invited Talk 504:25 PM - 05:00 PM

Lourdes Agapito

Personal Assistants

Lourdes Agapito
Bio

Lourdes Agapito is a professor at University College London (UCL), where she leads a research group working on 3D reconstruction from video. Her research focuses on the especially challenging task of estimating 3D shape of non-rigid objects. She recieved an ERC Starting Grant for her research contributions in 2008. In addition to her academic career, Prof. Agapito is a co-founder and director of Synthesia Inc., a startup empowering storytellers with AI.

05:00

Invited Talk 605:00 PM - 05:35 PM

Stefanos Zafeiriou

Collecting 3D Head Datasets

Stefanos Zafeiriou
Bio

Stefanos Zafeiriou is a Professor of Machine Learning and Computer Vision with the Department of Computing, Imperial College London. From 2016 to 2020, he was a Distinguishing Research Fellow with the University of Oulu, Finland, under Finish Distinguishing Professor Program. Prof. Zafeiriou is an EPSRC Early Career Research Fellow. He was a recipient of the Prestigious Junior Research Fellowships from Imperial College London in 2011, the President’s Medal for Excellence in Research Supervision for 2016, the President’s Medal for Entrepreneurship in 2022, the Google Faculty Research Awards, and the Amazon Web Services (AWS) Machine Learning (ML) Research Award.

05:35

Panel Discussion05:35 PM - 06:20 PM

None

Panel Discussion

Workshop Challenges

The workshop holds a competition on the newly introduced NeRSemble benchmark for 3D Head avatars. The goal is to find the current best method for dynamic novel view synthesis on heads and monocular FLAME-driven avatar reconstruction.

Dynamic Novel View Synthesis Challenge

Given synchronized multi-view videos from 13 cameras, the task is to replay the same facial action performance from the 3 hold-out viewpoints. This requires reconstructing a 4D representation that plausibly models both geometry and complex motion. The challenge is conducted on 5 short sequences from different individuals. The sequences cover complex dynamic effects such as topological changes when the tongue sticks out, flying hair, dynamic wrinkle changes, as well as light refraction and reflection at glasses. To capture even subtle movements, the sequences have been recorded at 73fps.

Monocular FLAME Avatar Challenge

Given several frontal videos of a person's head with corresponding tracked meshes from FLAME, the task is to re-animate the person with unseen FLAME expression codes and then render from both seen (blue) and unseen (orange) camera viewpoints. This requires reconstructing an animatable 3D head representation (=3D head avatar). The challenge is conducted on recordings from 5 different individuals. For each individual, 18 short facial performance sequences are provided for training while the remaining 4 sequences are hold-out. For the hold-out sequences, only the tracked FLAME meshes and the camera poses are known.

Competition Prizes

The winner of each workshop challenge will receive:
  • a dedicated 15-minute oral presentation in the workshop to showcase your method
  • an RTX 5080 GPU sponsored by NVIDIA*

*cannot be gifted to non-academics or persons residing outside of North America and Europe due to export restrictions imposed on NVIDIA by the US government

Competition Timeline

Date
Challenge begin 17th March 2025
Challenge submission deadline 23rd May 2025
Winner announcement 30th May 2025

Workshop Organizers

Tobias Kirschstein Tobias Kirschstein Technical University of Munich
Simon Giebenhain Simon Giebenhain Technical University of Munich
Tianye Li Tianye Li NVIDIA
Koki Nagano Koki Nagano NVIDIA
Justus Thieß Justus Thieß Technical University of Darmstadt
Matthias Nießner Matthias Nießner Technical University of Munich

Workshop Sponsors

NVIDIA We thank NVIDIA for sponsoring the prices for the workshop challenge winners.

Please contact Tobias Kirschstein for questions.