Accelerator Engine for Sound Localization Algorithm (CESLA)



  • Android
  • Java
  • Kotlin
  • Tensorflow Lite
  • Bluetooth



  • 2021


Accelerator Engine for Sound Localization Algorithm (CESLA) is a capstone project that aims to implement binaural sound localizaiton as a proof of concept using research done by Dr. Rong Zheng and Dr. Awny El-Mohandes of the McMaster Computing and Software department. The goal of this project is to receive external sounds via microphones, determine the direction of the sound based on machine learning models, and output the result to the user. An example use case for this system is for people whom are listening to music and may not be paying attention to dangerous traffic while crossing the road. The system will alert them if a sound appears to be dangerous.

What I worked on specifically during this project was obtaining data points for calibration of the machine learning algorithm, implementing classification of sounds, and Android application development.