Muhammed Kocabas
Machine Learning Researcher at Apple
United States
Summary
🤩
Rockstar🎓
Top SchoolMuhammed Kocabas is a machine learning researcher with a decade of experience specializing in 3D human pose and shape estimation, currently working at Apple and affiliated as a Guest Scientist at the Max Planck Institute for Intelligent Systems. He completed a PhD in Computer Science at ETH Zürich/Max Planck, and his work bridges academic research and industry productization through internships and roles at NVIDIA, Meshcapade, and Apple. Muhammed contributed to the widely used VIBE project (CVPR2020) on video-based human body pose and shape estimation, improving model outputs and integration of 3D joint/vertex data into inference pipelines. He combines deep expertise in SMPL-based modelling and practical engineering—frequently moving research prototypes toward robust, production-ready implementations.
10 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at Max Planck SocietyDoctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at ETH ZürichMaster of Science - MS Computer Engineering, Master of Science - MS Computer Engineering at Orta Doğu Teknik Üniversitesi / Middle East Technical UniversityBachelor of Science - BS Computer Engineering, Bachelor of Science - BS Computer Engineering at Istanbul Technical University
Github Skills (4)
human-pose-estimation10
pytorch10
computer-vision9
python9
Programming languages (6)
Github contributions (5)
mkocabas/VIBE
Dec 2019 - Aug 2022
Official implementation of CVPR2020 paper "VIBE: Video Inference for Human Body Pose and Shape Estimation"
Role in this project:
ML Engineer
ML EngineerContributions:2 releases, 2 reviews, 64 commits in 2 years 8 months
Contributions summary:Muhammed primarily focused on implementing and integrating 3D joint and vertex data (`joints3d`) within the VIBE model's inference and processing pipeline. This involved modifications to `demo.py`, `lib/utils/demo_utils.py`, and model-related files to incorporate and utilize the `joints3d` output during inference and SMPLify refinement. Additionally, the user made minor adjustments to data preparation scripts and conda/pip install instructions. The commits suggest improvements to the model's output and processing capabilities for 3D human pose and shape estimation.
inferencevideo-pose-estimation3d-pose-estimationhuman-pose-estimationsmpl
mkocabas/PARE
Oct 2021 - Apr 2024
Code for ICCV2021 paper PARE: Part Attention Regressor for 3D Human Body Estimation
Contributions:4 pushes, 10 comments, 10 issues in 2 years 6 months
3dhuman-pose-estimation3d-human-shape-and-pose-estimation3d-human-reconstructioncomputer-vision
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.