Ivan is currently unavailable
Ivan Makeev
Verified Expert in Engineering
Mathematics Developer
London, United Kingdom
Toptal member since March 5, 2014
Bio
Ivan is an algorithm developer and research scientist (PhD) with a strong background in mathematics and algorithms. He is interested in complex projects spanning various aspects of computer science, particularly deep learning, computer vision, and video analytics.
Portfolio
3DLOOK
Numerical Methods, Image Processing, Mathematics, Augmented Reality (AR)...
Humanoid
Algorithms, 3D Simulations, Diffusion Models...
Arrival
Image Processing, Mathematics, OpenCV, Deep Learning, Machine Learning...
Experience
- Mathematics - 15 years
- Computer Vision - 12 years
- Algorithms - 12 years
- Deep Learning - 10 years
- OpenCV - 10 years
- C++ - 10 years
- TensorFlow - 7 years
- Python - 6 years
Preferred Environment
Python, PyTorch, OpenCV
The most amazing...
...thing I've made is a foundational model for monocular 6DoF object pose estimation for flexible robotics assembly.
Work Experience
Chief Scientist
2016 - PRESENT
3DLOOK
- Owned the technology and R&D strategy and directions. Hired and led a team of seven engineers and researchers for core algorithmic R&D in human pose and body shape estimation.
- Developed one of the most accurate mobile 3D scanning solutions on the market.
- Developed neural architectures for high-precision computer vision problems.
- Built morphable parametric 3D models and energy-based optimization frameworks.
- Created tech stack for virtual try-on MVP based on generative AI using latent diffusion models.
Technologies: Numerical Methods, Image Processing, Mathematics, Augmented Reality (AR), OpenCV, Deep Learning, Machine Learning, Computer Vision, Algorithms, TensorFlow, Python, Artificial Intelligence (AI), Convolutional Neural Networks (CNNs), PyTorch, 3D Geometric Analysis, 3D Scanning, 3D Simulations, Diffusion-based AI Models, Scikit-learn, Bayesian Statistics, Statistics, Data Analysis, Data Structures, Large Language Models (LLMs), Image Generation, Data Science, Pandas
Research Engineer
2024 - 2025
Humanoid
- Developed learning-based robotic manipulation policies for bimanual robots.
- Developed imitation learning pipelines for robot control using deep generative models based on denoising diffusion, flow matching, and variational autoencoders.
- Developed a human-in-the-loop robot learning framework for fine-tuning of manipulation policies using offline reinforcement learning.
- Built tools for manipulation policies validation in simulated and hybrid environments.
Technologies: Algorithms, 3D Simulations, Diffusion Models, Generative Artificial Intelligence (GenAI), Robot Controls, Deep Reinforcement Learning, PyTorch, MuJoCo, Sensor Data, Data Analysis, Time Series Analysis, Data Structures, Large Language Models (LLMs), Reinforcement Learning, Data Science
Head of Deep Learning
2019 - 2024
Arrival
- Led applied algorithmic research and development in robotic perception and control.
- Authored and developed up to production quality a state-of-art learning-based algorithmic framework for monocular 6DoF pose estimation for robotic grasping.
- Optimized neural architectures for embedded platforms.
- Developed a reinforcement learning-based locomotion controller for a quadruped robot.
- Designed a learning-based computer vision pipeline for defects analysis for battery pack assembly lines.
Technologies: Image Processing, Mathematics, OpenCV, Deep Learning, Machine Learning, Computer Vision, Algorithms, C++, Python, TensorFlow, Artificial Intelligence (AI), Convolutional Neural Networks (CNNs), PyTorch, 3D Geometric Analysis, 3D Simulations, Scikit-learn, Sensor Data, Digital Signal Processing, Data Analysis, Time Series Analysis, Data Structures, Image Generation, Reinforcement Learning, Deep Reinforcement Learning, Data Science
Principal Research Engineer
2015 - 2019
V-Nova
- Conducted algorithms R&D for learning-based image and video compression.
- Authored and co-authored a number of international patents and contributed to technical standards.
- Implemented original image and video compression algorithms targeting real-time performance on heterogeneous hardware.
Technologies: Numerical Methods, Image Processing, Mathematics, C, Deep Learning, Machine Learning, Computer Vision, Algorithms, TensorFlow, Python, C++, Artificial Intelligence (AI), Convolutional Neural Networks (CNNs), NVIDIA CUDA, Sensor Data, Digital Signal Processing, Data Analysis, Time Series Analysis, Data Structures, Reinforcement Learning, Deep Reinforcement Learning, Data Science, Pandas
Research Scientist
2011 - 2015
Intel Labs
- Conducted research and development in image processing and computer vision.
- Investigated state-of-the-art algorithms and analyzed their applicability for real-world problems.
- Implemented proof-of-concept solutions.
- Disclosed my own ideas for patent applications.
- Analyzed third-party patents.
Technologies: Image Processing, Mathematics, Machine Learning, Computer Vision, Algorithms, OpenCV, C++, C, Convolutional Neural Networks (CNNs), Sensor Data, Statistics, Digital Signal Processing, Data Analysis, Data Structures, Data Science, Pandas
Associate Professor
2006 - 2015
Saint Petersburg State University
- Maintained research and development activities in computer vision, image processing, and automatic control.
- Participated in a number of joint international research projects.
- Taught graduate level courses in image processing, differential equations, and control theory.
Technologies: Numerical Methods, Image Processing, Mathematics, C, Computer Vision, C++, Data Analysis, Data Structures, Data Science
Senior Developer
2009 - 2011
Encoding.com
- Created a cross-platform desktop application (Adobe AIR) interfacing a cloud-based media encoding service via an API.
- Created a custom Flash video player (Vid.ly).
- Built a media uploader Flash app.
- Consulted back-end developers on encoding standards and codec settings.
Technologies: C, Algorithms, Digital Signal Processing
Algorithm Developer
2009 - 2011
Imagineer Systems
- Led algorithmic development for an Imagineer Mocha Pro, one of the most popular tools in the field of post-production and VFX, widely used by cinema artists.
- Developed a high-performance computational engine solving a number of computer vision problems.
- Integrated solutions with a variety of third-party software, including Adobe After Effects, Autodesk Maya, Maxon Cinema4D, and The Foundry Nuke.
Technologies: Image Processing, Mathematics, Augmented Reality (AR), OpenCV, C, Computer Vision, Algorithms, Python, C++
Senior Developer
2006 - 2010
Russian Web Service
- Created a platform solution for 3D modelers, including plugin utilities extending Autodesk 3DMax, render farm management tools, real-time data feed processors, 3D viewers, and more.
Technologies: C++
Experience
3DLook
https://3dlook.me/Body measurements using a mobile phone camera for fashion and fit tech, with deep deformable parametric models for human body pose and shape estimation and custom deep neural computer vision architectures. Research and development.
Observe Technologies
Deep recurrent fully-convolutional real-time architectures for simultaneous object detection, classification, and tracking for applications in aquaculture.
V-Nova
Deep learning for image and video compression. End-to-end trainable image compression architectures under challenging real-time constraints.
BorisFX (Imagineer) Mocha
https://borisfx.com/products/mocha/Post-production and VFX platform widely used by cinema artists. Implemented algorithms for precise object tracking and structure-from-motion.
Snappr Photo Analyzer
LinkedIn profile photo analyzer for Snappr.
Encoding.com
http://www.encoding.com/I created a cross-platform desktop app (Adobe AIR) interfacing the world's largest encoding service.
Education
2005 - 2008
PhD Degree in Applied Mathematics, Computer Science
Saint-Petersburg State University - Saint-Petersburg, Russia
2000 - 2005
Master's Degree in Applied Mathematics, Computer Science
Saint-Petersburg State University - Saint-Petersburg, Russia
Skills
Libraries/APIs
TensorFlow, OpenCV, PyTorch, Pandas, Scikit-learn
Tools
Robot Controls
Languages
C, Python, C++
Platforms
NVIDIA CUDA
Other
Algorithms, Mathematics, Computer Vision, Data Structures, Deep Learning, Sensor Data, Machine Learning, Image Processing, Artificial Intelligence (AI), Convolutional Neural Networks (CNNs), Image Generation, Data Science, Time Series Analysis, Numerical Methods, Data Analysis, Augmented Reality (AR), Statistics, Digital Signal Processing, 3D Scanning, Diffusion-based AI Models, Diffusion Models, Generative Artificial Intelligence (GenAI), Deep Reinforcement Learning, Large Language Models (LLMs), Reinforcement Learning, 3D Geometric Analysis, Bayesian Statistics, 3D Simulations, MuJoCo
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring