Computational Analysis and Acceleration Research Group (CARG)

Computational Analysis and Acceleration Research Group (CARG)

Biomedical and UAV related research

CARG HUB: A Versatile Innovation Center

HUB projects

CARG HUB serves as a versatile space that supports a wide range of projects, from individual student initiatives to large-scale collaborations with industry partners or government agencies. It places a strong emphasis on the diverse applications of machine learning and signal processing. CARG HUB is a dynamic environment that promotes interdisciplinary research and innovation.

Ongoing and/or recently completed projects at CARG HUB are:

  1. CARG HUB: Badminton stroke classification
    • Student/Researcher: Jeremy Cote
    • Status: Ongoing
    • This project presents an enhanced deep neural network for badminton stroke detection by leveraging both players’ movements and the shuttle trajectory.
  2. CARG HUB: AI for bioreactors
    • Collaborators: NRC, Ottawa; McGill University, Montreal
    • Status: Completed in 2023
    • This project aimed to leverage machine learning algorithms to predict cell growth in bioreactors, helping optimize bioprocesses for improved efficiency and yield.
  3. Notification System for Efficient Scheduling using Reinforcement Learning
    • Collaborators: Carleton University
    • Status: Completed in 2023
    • The system monitors a cluster of Digital Twins for critical states and sends notifications to predefined recipients within a strict deadline. It addresses challenges in scalability, cost-efficiency, resource utilization, and scheduling by using a micro-service-based architecture, multi-objective optimization, and reinforcement learning for efficient scheduling.
  4. AI Methods for Automated Software Testing
    • Collaborator: Ericsson, Ottawa
    • Status: Completed in 2020
    • This initiative focused on the use of artificial intelligence to create efficient, automated testing methods for software, minimizing human intervention while improving reliability.
  5. Federated Learning for Addressing Data Privacy Issues
    • Collaborator: IMRSV Data Labs, Ottawa
    • Status: Completed in 2020
    • This project explored federated learning as a way to address data privacy challenges, allowing models to be trained across decentralized devices without the need for centralized data storage.
  6. Multi-Microphone Signal Processing and Machine Learning
    • Collaborator: UCIC Toronto
    • Status: Completed in 2018
    • The project involved processing signals from multiple microphones using machine learning to enhance sound quality and recognition, contributing to advancements in communication and audio analysis systems.