YMYoushaa MurhijComputer Vision Engineer

Portfolio 2026

Youshaa Murhij

Computer Vision Engineer — Autonomous Driving

“I build perception systems that help machines understand the world.”

Work With MeDownload CV
Youshaa Murhij profile picture

About

About Me

I’m a Senior Computer Vision Engineer and ML Researcher (PhD in AI & ML) specializing in L4 autonomy, camera-based 3D perception, and sensor fusion. Over the past few years, I’ve developed and deployed end-to-end perception pipelines used in autonomous vehicle platforms and real-time simulation environments.

My focus is on translating deep learning research into reliable, production-ready systems — from multi-camera 3D scene understanding to embedded AI optimization.

Highlights

Key Achievements & Technical Strengths

Key Achievements
  • Delivered high-accuracy 3D object detection models for real-world driving scenarios.
  • Led integration of multi-sensor fusion pipelines (camera, LiDAR, radar).
  • Published peer-reviewed research and patented perception algorithms.
Deep Learning & CV
PyTorch
TorchVision
ONNX
OpenCV
NumPy
CUDA
TensorRT
Autonomous Driving
3D Vision
Multi-sensor Fusion
Occupancy Networks
Motion Prediction
Software Engineering
Python
C++
ROS
Bash
CI/CD
Docker
SQL

External Validation

8+

Peer-Reviewed Publications

2

Patented Software Components

CVPR, IJCNN, Neurocomputing

Top Venues & Journals

Timeline

Professional Experience

2023 - Present

Senior Computer Vision Engineer

Atom, Moscow, Russia

Responsible for developing and implementing advanced computer vision algorithms for autonomous systems.

  • - Lead 3D perception module development for L4 autonomous driving stack.
  • - Developed camera-based 3D detection and multi-view fusion models using PyTorch and CUDA.
  • - Optimized inference for embedded deployment on NVIDIA Jetson platforms.
  • - Collaborated with planning and control teams to improve perception-planning interface performance.

2022 - 2023

ML R&D Developer

Autonomics.Tech, Limassol, Cyprus

Focused on research and development of machine learning models for perception tasks in self-driving vehicles.

  • - Designed multi-modal sensor fusion systems for autonomous sweeper robot.
  • - Implemented and benchmarked object tracking algorithms for dynamic environments.
  • - Worked on real-time data pipelines for driving scenarios using C++, Python, and ROS.
  • - Supported field testing, dataset annotation automation, and perception stack integration.
  • - Assisted in developing software packages for obstacle detection and trajectory prediction.

2021 - 2023

Self-driving Perception Systems Developer

IntegraNT, Moscow, Russia

Developed perception systems for autonomous vehicles, including sensor fusion and object detection.

  • - Conducted research and implementation of computer vision and deep learning algorithms for intelligent robotic systems.
  • - Developed end-to-end object detection and segmentation networks for real-time applications in robotics.
  • - Implemented training pipelines and model optimization techniques for large-scale datasets using PyTorch.
  • - Contributed to publications on 3D object detection and feature map flow networks.
  • - Improved system inference efficiency by 30% through model pruning and CUDA-level optimization.

2020 - 2023

AI/CV Researcher

Cognitive Modeling Center, MIPT

Conducted research in artificial intelligence and computer vision, contributing to publications and projects.

  • - Conducted PhD-level research on AI perception systems for autonomous driving and robotics.
  • - Developed Lidar-based 3D object detection algorithms and occupancy-flow prediction models.
  • - Published multiple papers and patents in computer vision and 3D perception.
  • - Mentored graduate students and collaborated with research partners on applied AI projects.

Gallery

Publications & Patents

Explore peer-reviewed publications and patented software components through interactive case cards.

OFMPNet publication image

OFMPNet: Deep end-to-end model for occupancy and flow prediction in urban environment

Journal

Neurocomputing 2024

Authors

Youshaa Murhij and Dmitry Yudin

Project PagePaperCode
DAGM-Mono publication image

DAGM-Mono: Deformable Attention-Guided Modeling for Monocular 3D Reconstruction

Journal

Optical Memory and Neural Networks 2024

Authors

Youshaa Murhij and Dmitry Yudin

Project PagePaperCode
Abstract image for a patent on obstacle detection

Software package for detecting and tracking obstacles based on point cloud processing

Patent

Patent No. 2024613107 · Issued Feb 8, 2024

Authors

Youshaa Murhij

A real-time software package for detecting moving objects and plotting their trajectories from point clouds on embedded systems, designed for ground-based mobile robots.

View Patent
Abstract image for a patent on trajectory prediction

Software component for predicting the trajectory of dynamic obstacles

Patent

Patent No. 2023688827 · Issued Dec 25, 2023

Authors

Youshaa Murhij

A software component using graph neural networks to predict nonlinear trajectories of dynamic obstacles, integrated with detection systems via ROS.

View Patent
Feature Map Flow network visualization

FMFNet: Improve the 3D Object Detection and Tracking via Feature Map Flow

Conference

The International Joint Conference on Neural Networks (IJCNN) 2022

Authors

Youshaa Murhij and Dmitry Yudin

Project PageIJCNNPaperCode
Voxelization and Classification visualization

Rethinking Voxelization and Classification for 3D Object Detection

Conference

The 29th International Conference on Neural Information Processing (ICONIP) 2022

Authors

Youshaa Murhij and Dmitry Yudin

Project PageICONIPPaperCode
CVPR 2021 WAD Workshop

Real-time 3D Object Detection using Feature Map Flow

Conference

Computer Vision and Pattern Recognition (CVPR) 2021 WAD Workshop

Authors

Youshaa Murhij and Dmitry Yudin

Project PageCVPR 2021 WADPaperCode
VR application for robotics control

Towards faster response in virtual reality applications for robotics control

Journal

AIP Conference Proceedings 2022

Authors

Youshaa Murhij and Vladimir Serebrenny

AIP ProceedingsPaperCode
Hand gesture recognition for AR

Hand gestures recognition model for Augmented reality applications

Journal

The 15th International Conference on Electromechanics and Robotics 'Zavalishin's Readings'

Authors

Youshaa Murhij and Vladimir Serebrenny

Zavalishin's ReadingsPaperCode
Industrial robot simulation in VR

An application to simulate and control industrial robot in virtual reality environment integrated with IR stereo camera sensor

Conference

19th IFAC Conference on Technology, Culture and International Stability (TECIS) 2019

Authors

Youshaa Murhij and Vladimir Serebrenny

TECIS 2019PaperCode
Quadcopter indoor positioning system

Low-Cost Quadcopter Indoor Positioning System Based on Image Processing and Neural Networks

Conference

First International Congress for the Advancement of Mechanism, Machine, Robotics and Mechatronics Science (ICAMMRMS) 2017

Authors

I. Hatem, M. Jamal, Youshaa Murhij and Z. Ali

Paper

Academic

Education

2020 - 2024

Ph.D. in Computer science

Moscow Institute of Physics and Technology, Moscow, Russia

Advisor: Prof. Dmitry Yudin, Grade: 10/10

2018 - 2020

M.S. with honors in Mechatronics and Robotics

Bauman Moscow State Technical Universty, Moscow, Russia

Advisor: Prof. Vladimir Serebrenny, Grade: 5/5

2017 - 2018

Preparatory Faculty Student

Moscow State University of Geodesy and Cartography, Moscow, Russia

2012 - 2017

B.E. with honors in Mechatronics, Robotics and Automation Engineering

Grade: 89/100

Continuous Learning

Training & Courses

Contact

Get in Touch

Contact Information
Feel free to reach out via email or connect with me on social media. I'm always open to discussing new projects, creative ideas, or opportunities.
GitHubLinkedInX/TwitterORCIDResearchGateGoogle Scholar
Send a Message
Use the form below to send me a direct message.