Open to opportunities

Глеб Бобровских

@bobrg

Message

I build computer vision systems for 3D reconstruction, medical imaging, and robotics.

Russian Federation

Message

What I'm looking for

I'm looking to build impactful computer vision and deep learning products in 3D reconstruction, medical imaging, robotics, or generative AI, while collaborating with researchers, engineers, and domain experts.

I'm a Research Engineer at Skolkovo Institute, where I develop multi-modal 3D reconstruction models and LLM-based 3D generation from point clouds. My work improves reconstruction accuracy and completeness on textureless and reflective surfaces and has been reported in scientific publications.

I've also built medical-imaging systems for brain tumor segmentation from PET data and GNN-based MRI interpretability for disorder detection, collaborating closely with clinicians. In robotics perception, I integrated semantic priors into NeRF models for scene understanding.

At LiteBC, I developed blood-cell detection, tracking, trajectory extraction, velocity-estimation, and hemoglobin-prediction pipelines using YOLOv5, YOLOv8, Deep SORT, and VAEs. Earlier at Skolkovo, I helped collect a 12k+-frame multi-scanner dataset across 100+ scenes and 10+ lighting conditions for 3D reconstruction research.

My research spans 3D deep learning, geometry, depth super-resolution, object detection, and large-scale datasets, with publications at CVPR, SIGGRAPH, ICCV, and Doklady Mathematics. I also contributed to an agentic brain-health project, HelloMRI, through MRI analysis and rendering backend work.

Experience

Work history, roles, and key accomplishments

Current

Research Engineer

Current

Skolkovo Institute

Jun 2024 - Present (2 years 2 months)

Developed a multi-modal 3D reconstruction model with improved accuracy and completeness in textureless/reflective surfaces; implemented novel LLM-based 3D generation from point clouds. Collaborated with clinicians on brain tumor segmentation and applied GNN-based interpretability to MRI data.

3D Reconstruction 3D Generation Large Language Models NeRFs Medical Imaging 3D UNet GNN Python PyTorch

Middle Computer Vision Researcher

LiteBC

Feb 2023 - Jun 2024 (1 year 4 months)

Detected and tracked blood cells from capillary videos using YOLOv5, Deep SORT, and YOLOv8 models; implemented custom pipelines for blood flow trajectory extraction and cells velocity estimation. Collaborated with clinicians for clinical relevance of the research.

Computer Vision YOLO V5 YOLOv8 Python OpenCV Data Processing

Researcher

Skolkovo Institute

Nov 2018 - Jun 2024 (5 years 7 months)

Conducted 3D deep learning and computer vision research on range image super-resolution, 3D object detection, and learnable geometry, participating in two publications. Collected a large-scale dataset with 3D scanners and processed data from SUNRGBD, Middlebury, and ABC datasets.

Computer Vision Range Image Super Resolution 3D Detection Data Acquisition Python PyTorch

Data Science Intern

Mobile TeleSystems

Jun 2018 - Oct 2018 (4 months)

Conducted natural language processing research on intellectual QA systems, named entity recognition, and multi-label classification. Applied open-source methods to novel data and performed data mining and analysis.

Natural Language Processing Named Entity Recognition Data Mining Python

Teaching Assistant

Skolkovo Institute

Jan 2018 - Jul 2018 (6 months)

Examined homeworks completed by a group of 30+ students across 5 problem sets and answered questions related to course materials. Course topics included neural network basics, backpropagation, optimizers, and CNN architectures.

Teaching Neural Networks Backpropagation Optimization Convolutional Neural Networks