Topics and areas of my expertise

• Research & development
  • real-time deep learning models
    • efficient neural networks for dense per-pixel tasks
  • deployment of deep learning models
    • cloud / server
    • in-browser
    • mobile app
    • embedded devices
  • data labeling
    • scoping data labelling costs
    • evaluating feasibility of using real vs synthetic data
  • setting up end-to-end reproducible pipelines
    • cloud or on-prem
    • from raw data extraction to a deployed model
  • fusion of deep learning with classical computer vision and geometrical approaches
    • solving real-world robotics’ perception problems
• Programming
  • scientific computing, data science in Python and R
  • distributed approaches for data processing
  • interactive visualisation, dashboard tools
  • machine learning, deep learning, computer vision solutions from scratch
  • deployment of deep learning models with C++ and CUDA