Kfir Aberman
Distinguished Member at Decart
San Francisco Bay Area United States
Summary
🤩
Rockstar🎓
Top SchoolKfir Aberman is a research-driven AI leader with nine years of experience building cutting-edge computer-graphics and multimodal generative systems, currently founding Decart’s US effort to create emotionally intelligent, real-time audio-visual agents. He has led research and product teams at Snap and Google, shipping personalized generative features for consumer devices and guiding high-impact ML engineering. His background spans academia and industry—from PhD work to a SIGGRAPH-linked open-source contribution in deep motion editing that bridges deep learning and Blender integration—to applied R&D in reality capture and visual effects. Comfortable moving projects from core model design to optimized inference stacks, he combines hands-on engineering, team leadership, and a history of deploying research into production-grade experiences.
9 years of coding experience
12 years of employment as a software developer
Master of Science (M.Sc.) Electrical Engineering, Master of Science (M.Sc.) Electrical Engineering at Technion - Israel Institute of TechnologyDoctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Tel Aviv University
Hebrew, English, Spanish, Chinese
Github Skills (10)
deep-learning10
computer-graphics10
python10
style-transfer9
machine-learning9
pytorch8
animate8
animation7
3d7
blender7
Programming languages (2)
Github contributions (5)
DeepMotionEditing/deep-motion-editing
Apr 2020 - Aug 2020
An end-to-end library for editing and rendering motion of 3D characters with deep learning [SIGGRAPH 2020]
Role in this project:
ML Engineer
ML EngineerContributions:35 commits, 20 PRs, 26 pushes in 3 months
Contributions summary:Kfir appears to be involved in developing and structuring components for a deep learning-based motion editing library. They added the core structure of the project with a focus on options and testing. The user also added code related to model creation, suggesting they're working with different motion editing models such as retargeting and style transfer. Furthermore, the user added files for Blender integration, adding joint rotations and translations.
style-transferend-to-endcomputer-animationdeep-learningoptical-flow
DeepMotionEditing/DeepMotionEditing.github.io
May 2020 - Jun 2020
A webpage for the papers "Skeleton-Aware Networks for Deep Motion Retargeting" and "Unpaired Motion Style Transfer from Video to Animation" - SIGGRAPH 2020
Contributions:39 commits, 38 PRs, 18 pushes in 1 month
animationstyle-transferretargetingskeletonmotion
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.