Skip to content

Johnson Loh

Biography

My professional passion is to improve general human health and well-being through technological solutions. As a generalist with many interests, I like to challenge myself in new application fields with a recent focus on bio-inspired computing in biomedical signal monitoring and I am looking forward to my next career challenge. I thrive most in multi-disciplinary projects and I like to solve them in a multi-cultural harmonious team. I strongly believe that the creativity and high ethical standards are best fostered in a diverse team and are necessary to solve the challenges of the future.

Project Summary PhD
Project Summary Master
  • Master studies at RWTH Aachen University in Electrical Engineering, Information Technology and Computer Engineering
  • Simultaneous Localization and Mapping (SLAM) and basic navigation for SeekurJr robot in VEROSIM (Master's Thesis at Institute for Man-Machine Interaction) (Thesis (ger))
    • Seekur Jr is equipped with 3 front, 1 rear camera (calibrated/interfaced with OpenCV) and one front LIDAR
    • Extended and Unscented Kalman Filter with extensible state variable for SLAM
    • SLAM map consists of hybrid grid and feature map in 2D space
    • Navigation is realized with global gridmap-based path planning and local collision avoidance
    • Implementation is integrated into proprietary VEROSIM environment (custom micro kernel architecture) with numerical computing libraries, e.g. Eigen library
  • Integration of SLAM features for existing ROS framework (Internship at Fraunhofer IPA group, now: NODE Robotics) (Report (ger))
    • Implementation of adaptive partical filter with KLD sampling
    • Implementation of GPS and RFID sensor modules with modular integration into ROS framework
      • Sensor fusion through particle filter algorithm in SLAM implementation
    • Implementation of map conversion algorithms between low-resolution normal distribution transform (NDT) and high-resolution gridmaps
      • Conversion algorithm is partially based on classifical computer vision methods, e.g. Bresenham adapted to ellipsoids or wavefront algorithms
  • Implementation of SLAM and navigation features in PenguiPi robot (Exchange semester in Queensland University of Technology)
    • Beacon-based localization detected through commercially available camera modules
    • Image pre-processing and noise reduction for reliable beacon detection in distance
    • Motor encoder and actuation calibration for reliable navigation in unknown environment
    • (Backup Repo)
  • R&D support for cardiac support systems (Student research assistant at Institute of Automatic Control)
    • Prototype development of printed circuit board (PCB) designs for sensor read-out and actuation
    • Embedded programming of microcontrollers in C/C++ and basic system modeling using standardized open-source specifications like functional mock-up interface (FMI)
Project Summary Bachelor

Disseminations

Publications (Google Scholar)

  • J. Loh, L. Dudchenko, J. Viga and T. Gemmeke, "Towards Hardware Supported Domain Generalization in DNN-based Edge Computing Devices for Health Monitoring," IEEE Transactions on Biomedical Circuits and Systems, Jun. 2024, doi: 10.1109/TBCAS.2024.3418085.

  • M. Lahr, J. Loh, M. Schwarz, M.Lemme, T. Gemmeke, "Vehicle Surroundings Perception using MEMS Inertial Sensors," Advanced Intelligent Systems, Apr. 2024, doi: 10.1002/aisy.202300679.

  • A. Ayad, M. Barhoush, J. Loh, T. Gemmeke, and A. Schmeink, "PEACE: Private and energy-efficient algorithm for cardiac evaluation on the edge using modified split learning and model quantization," in 2023 14th International Conference on Information and Communication Systems (ICICS), IEEE, Nov. 2023. doi: 10.1109/icics60529.2023.10330472.

  • J. Loh and T. Gemmeke, "Stream processing architectures for continuous ecg monitoring using subsampling-based classifiers," IEEE Transactions on Very Large Scale Integration (VLSI) Systems, pp. 1–11, 2023, ISSN: 1557-9999. doi: 10.1109/tvlsi.2023.3329360.

  • Y. Chen, J. Lou, C. Lanius, F. Freye, J. Loh, and T. Gemmeke, "An energy-efficient and area-efficient depthwise separable convolution accelerator with minimal on-chip memory access," in 2023 IFIP/IEEE 31st International Conference on Very Large Scale Integration (VLSI-SoC), IEEE, Oct. 2023. DOI: 10.1109/vlsi-soc57769.2023.10321918.

  • C. Lanius, J. Lou, J. Loh, and T. Gemmeke, "Automatic generation of structured macros using standard cells – application to CIM," in 2023 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), IEEE, Aug. 2023. DOI: 10.1109/islped58423.2023.10244608.

  • J. Loh and T. Gemmeke, "Lossless Sparse Temporal Coding for SNN-based Classification of Time-Continuous Signals," 2023 Design, Automation & Test in Europe Conference & Exhibition (DATE), Antwerp, Belgium, 2023, pp. 1-6, doi: 10.23919/DATE56975.2023.10137112.

  • J. Loh and T. Gemmeke, "Dataflow optimizations in a sub-uW data-driven TCN accelerator for continuous ECG monitoring," in 2022 IEEE Nordic Circuits and Systems Conference (NorCAS), IEEE, Oct. 2022. DOI: 10.1109/norcas57515.2022.9934591.

  • C. Latotzke, J. Loh, and T. Gemmeke, "Cascaded classifier for pareto-optimal accuracy-cost trade-off using off-the-shelf ANNs," in Machine Learning, Optimization, and Data Science, Springer International Publishing, 2022, pp. 423–435. DOI: 10.1007/978-3-030-95470-3_32.

  • J. Loh, J. Wen and T. Gemmeke, "Low-Cost DNN Hardware Accelerator for Wearable, High-Quality Cardiac Arrythmia Detection," 2020 IEEE 31st International Conference on Application-specific Systems, Architectures and Processors (ASAP), Manchester, UK, 2020, pp. 213-216, doi: 10.1109/ASAP49362.2020.00042.

Patent Applications

  • M. Lahr, R. Foell, T. Gemmeke and J. Loh, "PKW-Umgebungswahrnehmung mit MEMS Inertialsensoren und digitaler Auswertelogik," German Patent and Trade Mark Office, submitted Aug. 4, 2023.

  • T. Gemmeke, J. Loh, C. Höffler, A. Schmeink and A. Ayad, "Energieoptimierte Vorrichtung zur Klassifizierung von Daten," German Patent and Trade Mark Office, Patent No. DE102020202391A1, published Aug. 26, 2021, (discontinued Jan. 15, 2022).

Teaching

  • Supervised Students

    • Jianan Wen (M.Sc. - Thesis), 2020 - Logic design of an Energy-Efficient CNN for ECG Classification based on Wavelet Transformation
    • Qiwei Zhang (M.Sc. - Thesis), 2020 - Automated Design Space Exploration for Hardware Implementations of Structured Datapaths
    • Junhong Li (M.Sc. - Thesis), 2020 - Exploration of Hardware Architectures for ECG Classification using ResNet Architectures
    • Louis Cherel (M.Sc. - Thesis), 2021 - Hardware Architectures for 1D Sparsely Encoded Neural Networks
    • Maxim Drobjazko (M.Sc. - Thesis), 2022 - Recurrent NN Hardware-Architectures for Time-Continuous Biomedical Signals
    • Justus Viga (M.Sc. - Thesis), 2022 - In-Field Individualisation of Classification-/Detectionsystems for Biomedical Signals
    • Adeola Adebayo (M.Sc. - Thesis), 2023 - Hardware-Supported On-the-fly Domain Generalization for ECG Classification
    • Additional 6 student research assistants (+ 4 co-supervised)

  • Lectures and Labs

    • Supervision of
      • Seminar on Selected Topics of Integrated Digital Systems - "Neuromorphic Architectures - A Feasibility Study for Real-Time Edge Computing", RWTH Aachen University (M.Sc.)
      • Computer Arithmetic - Fundamentals, RWTH Aachen University (M.Sc.)
      • VLSI Design for Digital Signal Processing - Architectures, RWTH Aachen University (M.Sc.)
      • VLSI Design Lab, RWTH Aachen University (M.Sc.)
      • FPGA Design Lab, RWTH Aachen University (M.Sc.)
      • Micro-/Nanoelectronics Lab - VLSI and FPGA Design Fundamentals, RWTH Aachen University (B.Sc.)
    • Additional support for
      • Computer Arithmetic - Advanced, RWTH Aachen University (M.Sc.)
      • Artificial Neural Networks, RWTH Aachen University (M.Sc.)
    • (Website)

Services

Miscellaneous

Languages

  • German: Mother tongue (spoken & written)
  • English: Fluent (spoken & written)
  • Cantonese: Mother tongue (spoken)
  • Mandarin: Basic (spoken & written)

Extra-curricular Activities