Career Profile

I am a doctoral researcher at the Chair of Digital Signal Processing and Circuit Technology at TU Chemnitz, having joined after graduating from TU Dresden. My research focuses on applying computer vision and machine learning technologies to various industrial and social domains, including rehabilitation, active assisted living, autonomous driving, and railway safety. I actively participate in teaching activities, instructing courses on basic and advanced computer vision, 3D vision, Python for data scientists, and deep learning for computer vision. My Ph.D. thesis explores real-time human detection, pose estimation, and activity recognition using top-view omnidirectional imaging with fisheye cameras.

Experiences

Doctoral Researcher

July 2017 - Present
TU Chemnitz, Germany

The topic of my doctoral thesis is “Realtime Person Detection, Pose Estimation and Human Activity Recognition in Indoor Omnidirectional Scenes with Deep Neural Networks”. In addition to algorithm development, an important aspect of the topic is data generation, especially synthetic data generation using game engine (Unity), NeRF and 3DGS. Besides my doctoral research, I am engaged in other research activities:

  • Scientific projects with the focus on computer vision and machine learning.
  • Research funding application.
  • Student research thesis instruction.

Teaching Assistant

July 2017 - Present
TU Chemnitz, Germany
  • Computer Vision I: lecturing chapters on machine learning and tracking, guiding assignments, exam evaluation.
  • Computer Vision II: lecturing chapters on deep learning, camera geometry and 3D vision, guiding assignments, exam evaluation.
  • Programming and Data Analysis: co-designing module structure, lecturing on python language, guiding assignments and projects, exam evaluation.

Research Intern

May 2014 - November 2014
Fraunhofer IKTS, Dresden Germany
  • Conceptual design of a polymer reaction system

Projects

AuBBI - railway crossing surveillance with IR camera. 2024
SWARM - federated learning for autonomous driving platforms. 2024
ReViCaT - reversing camera for freight wagon coupling with precise distance measurement. 2023 - 2024
SPREAD - synthetic image and depth map generation with Unity for energy efficient and safe automated doors. 2023 - 2024
AISafe - synthetic image and depth map generation with Unity for the navigation of automated guided vehicle (AGV). 2022 - 2024
ReFit - AI assisted rehabilitation at home with real-time feedback. 2019 - 2021
KIN-TUC - real-time human pose estimation with deep learning. 2019 - 2021
AUXILIA - real-time person detection and human pose estimation with deep learning for indoor top-view fisheye camera in ambient assisted living. 2017 - 2021

Publications

  • NToP: NeRF-Powered Large-scale Dataset Generation for 2D and 3D Human Pose Estimation in Top-View Fisheye Images
  • Jingrui Yu, Dipankar Nandi, Roman Seidel, Gangolf Hirtz
    arXiv preprint arXiv:2402.18196. 2024
  • Human Pose Estimation in Monocular Omnidirectional Top-View Images
  • Jingrui Yu, Tobias Scheck, Roman Seidel, Yukti Adya, Dipankar Nandi, Gangolf Hirtz
    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023
  • Applications of Deep Learning for Top-View Omnidirectional Imaging: A Survey
  • Jingrui Yu, Ana Cecilia Perez Grassi, Gangolf Hirtz
    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023
  • OmniPD: One-step person detection in top-view omnidirectional indoor scenes
  • Jingrui Yu, Roman Seidel, Gangolf Hirtz
    Current Directions in Biomedical Engineering 5.1 (2019): 239-244
  • Contactless Interactive Fall Detection and Sleep Quality Estimation for Supporting Elderly with Incipient Dementia
  • Roman Siedel, Tobias Scheck, Ana C Perez Grassi, Julian B Seuffert, André Apitzsch, Jingrui Yu, Norbert Nestler, Danny Heinz, Lars Lehmann, Anne Goy, Gangolf Hirtz
    Current Directions in Biomedical Engineering 6.3 (2020): 388-391
  • Auxilia: Nutzerzentriertes assistenz-und sicherheitssystem zur unterstützung von menschen mit demenz auf basis intelligenter verhaltensanalyse
  • Roman Seidel, André Apitzsch, Jingrui Yu, Julian Seuffert, Norbert Nestler, Danny Heinz, Anne Goy, Gangolf Hirtz
    aw&I Conference, 2018

    Skills & Proficiency

    Python

    PyTorch

    Unity

    Linux