ETRO VUB
About ETRO  |  News  |  Events  |  Vacancies  |  Contact  
Home Research Education Industry Publications About ETRO

Master theses

Current and past ideas and concepts for Master Theses.

Embedding Deep Neural Networks for Urban Sound Recognition

Subject

Deep Neural Networks (DNN) can be used to recognize particular sounds. Such a solution demands an architecture exploration before being embedded in order to satisfy real-time and power-efficiency constraints. The ultimate goal is to combine the most promising DNN with an MEMS microphone array, which is used to determine the sound’s direction of arrival. Their combination would ultimately have the potential to identify acoustic patterns while determining the direction of arrival.

Kind of work

Work to be done:
1. Literature study of DNN-based solutions.
2. Develop and evaluate your own DNN architecture.
3. Compare the DNN-based solution with traditional Machine Learning approaches.
4. Embed and evaluate your architecture.

Framework of the Thesis

Embedded AI Techniques for Industrial Applications (AITIA) project
http://aitiaproject.eu/

Number of Students

1

Expected Student Profile

• Knowledge/Interest in Deep Neural Networks (DNNs).
• Good C/C++ or Python programming experience.

Promotors

Mr. Abdellah Touhafi

+32 (0)2 629 3774

atouhafi@etrovub.be

more info

Prof. Bruno da Silva Gomes

+32 (0)2 629 3768

bdasilva@etrovub.be

more info

- Contact person

- IRIS

- AVSP

- LAMI

- Contact person

- Thesis proposals

- ETRO Courses

- Contact person

- Spin-offs

- Know How

- Journals

- Conferences

- Books

- Vacancies

- News

- Events

- Press

Contact

ETRO Department

Tel: +32 2 629 29 30

©2022 • Vrije Universiteit Brussel • ETRO Dept. • Pleinlaan 2 • 1050 Brussels • Tel: +32 2 629 2930 (secretariat) • Fax: +32 2 629 2883 • WebmasterDisclaimer