Internship (Project - Data Collection and Analysis for AI-based ANC system) at HARMAN International
Kvistgård, , Denmark -
Full Time


Start Date

Immediate

Expiry Date

04 Dec, 25

Salary

0.0

Posted On

06 Sep, 25

Experience

0 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

No

Skills

Python, English, Matlab, Signal Processing, Communication Skills, Mathematics

Industry

Information Technology/IT

Description

Location:
Kvistgard - Denmark
Job Family:
General & Administration
Worker Type Reference:
Intern (Fixed Term) (Trainee) - Intern
Pay Rate Type:
Salary
Career Level:
T0
Job ID:
R-46039-2025

WE ARE LOOKING FOR A MOTIVATED INDIVIDUAL WITH THE FOLLOWING QUALIFICATIONS.

  • Master’s student with good grades in audio signal processing and mathematics.
  • Strong programming skill in MATLAB and Python.
  • Experience with using audio hardware and audio laboratories.
  • Excellent communication skills.
  • Excellent written and spoken English, ability to work as part of an international team.
Responsibilities

ABOUT THE ROLE:

The global New Technology & Innovation (NT&I) team is at the center of audio technology support in Harman’s innovative products. The team is seeking a talented and motivated student to join us for a six-month project on Data Collection & Analysis for AI-based ANC System.

WHAT YOU WILL DO:

Modern ANC systems are commonly based on using adaptive filters to react to variations in noise characteristics, wearing conditions and environments. We are now advancing to explore how we can use AI in ANC systems and to enable that, a database is needed. The scope of this project is primarily to build the database that can be used for exploring the AI solutions.
The main task of this project is to build a database of ANC related recordings and transfer functions measured on at least 20 people, and on multiple formfactors.
1. Perform recordings of different noise types in multiple conditions, using headphone prototypes equipped with transducers.
2. Analyze the data, deriving the relevant transfer functions and performing statistics.
3. Organize the dataset for convenient use.
4. Analyze whether the secondary path can be parameterized, and whether this is sensitive in terms of formfactor.
5. Hands-on ANC tuning based on the data gathered.
6. Additional work: Support in setting up automated tests for adaptive ANC.

Loading...