Research

Our skills

  • We have long experience what kind of sensors are needed to measure different signals and how to pre-process these signals for feature extraction and prepare it for training and validation.
  • We have long experience how to design embedded hardware and develop software for it. 
    • Long experience and collaboration with other partners regarding embedded hardware design.
    • Embedded software written for countless number of projects.
  • We know how to choose optimal machine learning models for embedded hardware and optimize it to reduce computational effort, of reduze the size for devices with small memory.
    • Our researchers are daily-basis working on such issues in different projects
  • We can build a new machine learning model or re-train the existing model in case you need to work with completely new kind of data with the help of machine learning.
  • We can help you to choose a right hardware accelerator to speed up computationally intensive machine learning models.
    • We have a wide range of hardware available to test different approaches.

Our research focus

Our research areas are divided into three main groups. They are all connected together to focus on novel optimization methods and prototyping to validate new approaches.

  • Machine Learning optimization techniques
    • Main competences
      • Development of new ML optimization technologies
      • Acceleration of ML inference with hardware accelerators
      • Development of ML models on FPGA
      • Image processing and compression
    • Side competences
      • Sensor signal measurement
      • Signal processing and feature extraction
      • Sensor fusion
  • Embedded System Design
    • Embedded hardware design
    • Embedded software design
    • Rapid prototyping