Jekaterina Jaroslavceva
Research Scientist (Computer Vision) at Microsoft
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolJekaterina Jaroslavceva is a research scientist in computer vision and machine learning with nine years of experience bridging cutting-edge research and production-grade systems. Currently on Microsoft’s Spatial AI team in Zurich, she previously developed a GNN-based 3D energy-deposit linking algorithm at CERN that outperformed rule-based benchmarks for HL-LHC calorimeter reconstruction. Her background includes creating a state-of-the-art CNN image retrieval method now integrated into the TensorFlow models repository and open-source contributions implementing contrastive and triplet losses and advanced pooling layers. She combines strong academic training in cybernetics and robotics with hands-on skills in Python, PyTorch, TensorFlow, C/C++ and ROS2, and has practical experience in robotics DSP and speech-noise suppression. Jekaterina balances rigorous R&D with creative side projects—award-winning AR and sign-language hardware—that reflect a practical, experiment-driven approach to ML. Colleagues describe her as a fast learner who thrives on complex, cross-disciplinary problems and collaborative teams.
9 years of coding experience
4 years of employment as a software developer
Bachelor's degree, Cybernetics and Robotics, Bachelor's degree, Cybernetics and Robotics at University of AdelaideGymnázium Jana KepleraHong Kong University of Science and Technology (HKUST)Master's degree, Cybernetics and Robotics, GPA 100%, Master's degree, Cybernetics and Robotics, GPA 100% at Faculty of Electrical Engineering, Czech Technical University in PragueBachelor's degree, Artificial Intelligence, Bachelor's degree, Artificial Intelligence at prg.ai Minor
English, Czech, Russian
Github Skills (8)
computer-vision10
machine-learning10
deep-learning10
tensorflow10
python10
modeling9
trainings9
keras8
Programming languages (3)
Github contributions (5)
tensorflow/models
Feb 2021 - Nov 2021
Models and examples built with TensorFlow
Role in this project:
ML Engineer
ML EngineerContributions:17 reviews, 8 commits, 9 PRs in 9 months
Contributions summary:Jekaterina's commits primarily focus on implementing and modifying ranking losses, normalization layers, and pooling layers within the TensorFlow models repository. The code changes reveal the addition of contrastive and triplet loss functions, as well as various pooling layer implementations such as MAC, SPoC, and GeM. The modifications to dataset utilities and model definitions suggest involvement in setting up and training these models, demonstrating expertise in building and adapting deep learning components.
deep-learningtensorflow
ykate1998/models
Mar 2021 - Nov 2021
Models and examples built with TensorFlow
Contributions:74 commits, 2 PRs, 69 pushes in 8 months
tensorflow
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.