The University of Auckland

Project #87: Capturing speech audibility in classrooms through machine learning techniques

Back

Description:

 

A successful learning experience requires students to be able to hear what the teacher is saying. Poor audibility in New Zealand classrooms is a problem for many students, potentially affecting their learning. Students need a high signal-to-noise ratio (SNR) or speech audibility to hear the teacher against background noise. This is especially critical for Students with hearing difficulties, and Students for whom English is not their first language.

The project will develop a system from previous work that utilises machine learning techniques to measure the SNRs and acoustics that affect speech comprehension in various listening positions in a typical classroom. The system will then be tested in classrooms such as lecture theatre and laboratories.

Students suitable for this project should be interested in education, acoustics, and software and hardware development. They should also have competent software skills. 

 

 

 

 

Type:

Undergraduate

Outcome:

Prerequisites

As this project requires fundamental knowledge in digital signal processing, at least one of the group members must have passed MECHENG370 Electronics and Signal Processing (i.e. the team must have at least one Mechatronics student).

Taking MECHENG726: Acoustics for Engineers (Part4 elective offered in S2) would help students understand the theories and techniques useful for the project (enrolment is not a requirement).

Specialisations

Categories

Supervisor

Team

Lab

No lab has been assigned to this project