Topic status: In progress
We have collected multiple terabytes of environmental sound using sensor networks. Some sounds have also been analysed either by algorithms or by human and corresponding tags have been generated. However, the relationships between different sounds and associated tags and other meta data such as location and time have not been analysed systematically.
This project involves gathering related literature, reading and critically analysing current literature and prepare a comprehensive literature review for this topic in publishable format. It will also involve some coding for proof-of-concept of an existing or innovative idea.
The expected outcomes are:
- a 20 page comprehensive literature review for the state-of-the-art of data mining techniques with application in environmental sound analysis
- a demo for a chosen idea.
Duration of project
8-10 weeks during summer semester.