VIP Projects: IC CAE Team
Goals
The purpose of the IC CAE Team is to investigate the application of state-of-the-art image recognition / image segmentation machine learning algorithms with Hume’s work in radio frequency machine learning (RFML) applications. The team will implement state-of-the-art approaches (such as the well-known YOLO algorithm) and determine their strengths and weaknesses for wireless spectrum sensing applications. Through this investigation, a novel hybrid approach leveraging the individual strengths of the approaches, will mitigating their independent weaknesses, will be developed and its performance evaluated.
Issues Involved or Addressed
- Applying state-of-the-art image recognition and image segmentation machine learning techniques to radio frequency applications
- Investigating the tradeoffs between image-based machine learning (image inputs) and radio frequency machine learning (rf data inputs)
- Develop collaborative image and rf based machine learning techniques that leverage the strengths of both approaches and mitigate their weaknesses
Methods and Technologies
- Python (code development and dataset generation)
- GNU Radio Companion (over-the-air testing)
- PyTorch (deep learning training, validation, and testing)
Academic Majors of Interest
- Computer Science
- Computer Engineering
- Computational Modeling and Data Analytics
- Electrical and Computer Engineering
- Other students with similar experience seeking future developer roles
Preferred Interests and Preparation
- Wireless Communications
- Dataset Creation
- Machine Learning
- Image Processing/Recognition