dr inż. Michał Kępski

  • Jednostka:
    Wydział Nauk Ścisłych i Technicznych, Instytut Informatyki
  • Budynek: A0 (korytarz B1), ul. prof. Stanisława Pigonia 1
  • Pokój: 351
  • Nr telefonu: 17 851 8599
  • e-mail: [email protected]
  • ORCID: 0000-0003-1225-9143
  • Konsultacje dla studentów: 11:30-13:00 poniedziałek (A) / wtorek (B)

Informacje

Field of interest

computer vision, image processing, machine learning, computer graphics

Grants

Project DUST – Development of Tools and Production Processes for Creating Off-Road Driving Video Games Using Unreal Engine 5
Role: R&D Manager
Organization: Simteract SA
Period: 2024–present
Funding: “European Funds for a Modern Economy” Operational Programme, “SMART Path” (FENG.01.01-IP.02-3931/23)

Generation of Realistic Visualization of Train Routes for Professional Railway Simulators
Role: R&D Manager
Organization: Simteract SA
Period: 2020–2022
Summary: Technology for data-driven procedural generation of 3D content for simulators and games was developed.
Funding: National Centre for Research & Development (POIR.01.01.01-000382/20)

A Fall Detection System for the Elderly Based on Fuzzy Logic and Machine Learning
Role: Principal Investigator
Organization: University of Rzeszow
Period: 2021–2022
Summary: An action recognition system, with emphasis on fall detection, was proposed. The solution used data fusion from depth cameras and IMUs.
Funding: Podkarpackie Innovation Centre (09/UR/1/DG/PCI/2020)

Development of a Dedicated Software Prototype of a Zone Sensor Based on Computer Vision and ToF Cameras
Role: Principal Investigator
Organization: University of Rzeszow
Period: 2018
Summary: Embedded software for patient-zone event recognition from 3D data was developed. Real-time activity detection, such as zone entry and person-to-person contact, was achieved using a deep CNN running on the NVIDIA Jetson TX2.
Funding: Regional Operational Programme (RPMP)

Development of a Dedicated Software Prototype of an Input/Output Sensor
Role: Principal Investigator
Organization: University of Rzeszow
Period: 2018
Summary: Embedded software for counting people from 3D data was developed. Real-time detection was achieved using a deep CNN on the NVIDIA Jetson TX2.
Funding: Regional Operational Programme (RPMP)

Awards and Distinctions
  • Best Paper Finalist, IEEE WCCI 2020,
  • Rector’s Award, 2017,
  • Best Presentation Award in the domain of Computer Vision, International Conference on Computer Vision and Graphics, ICCVG 2012.
Courses Taught (past and present)
  • Computer Graphics and Human-Computer Interaction,
  • Python Programming,
  • Image Processing and Analysis,
  • Introduction to C Programming,
  • Machine Perception and Learning.

Citation metrics (April 2026, Google Scholar): 1,600+ citations, h-index 14.

Publikacje