Master Thesis - Sound Quality Perception Research [f/m/n]

Join the leader in entertainment innovation and help us design the future. At Dolby, science meets art, and high tech means more than computer code. As a member of the Dolby team, you’ll see and hear the results of your work everywhere, from movie theaters to smartphones. We continue to revolutionize how people create, deliver, and enjoy entertainment worldwide. To do that, we need the absolute best talent. We’re big enough to give you all the resources you need, and small enough so you can make a real difference and earn recognition for your work.

Our company philosophy encourages creativity, collaboration, and the desire to perceive things differently.

At our Sound Tech Research Group in Nuremberg, we are looking to hire a student working towards finishing his/her Master Degree with a Thesis on modeling audio quality. In return, we offer working on a cutting-edge audio and video technologies, as well as lots of opportunities to network and collaborate with world-class researchers across the company.

Duration: 6 months

The position is based at our office in Nuremberg, Germany.

Essential Job Functions:

Estimating the quality of coded audio is a non-trivial task, particularly when parametric coding tools are employed. While traditionally listening test participants are asked to rate the quality of coded audio, e.g., in a MUSHRA test setup; automated methods are also available. Recently, deep learning-based models are shown to outperform signal processing-based models in predicting coded audio quality. However, such deep learning-based models are usually trained for a listening playback scenario where a large amount of data is available. For example, our deep learning-based coded audio quality prediction model was trained with subjective data where the listening tests were conducted in a controlled listening room environment and auditioned over headphones. If the listening environment is assumed to be the same, subjects may be more sensitive to certain coding artifacts audible over headphones but not over loudspeakers, and vice-versa. Thus, the overall quality achieved with either headphone or loudspeaker-based listening test may differ. However, there is a lack of sufficient subjective data (e.g., same listening test auditioned with both headphones and loudspeaker) as well as literature to draw a meaningful conclusion to verify this hypothesis and to possibly derive a mapping of quality from headphone to loudspeaker listening.

The goal of this Master Thesis is to take steps towards the development of a playback device-dependent quality model, specifically, investigate and decide if there is a need to develop a loudspeaker version of our quality model assuming listening in a listening room-like environment. The student is expected to perform literature research, conduct subjective tests from which we could derive any possible relationship in perceived quality for headphones versus loudspeaker playback, study the performance of our headphone-based quality model for loudspeaker listening test and propose a suitable mapping if needed.

Qualifications:

• Experience in Python/ MATLAB
• Experience in PyTorch
• Strong passion for audio quality.

Eligibility:

• Master students in multimedia technologies, machine learning, statistics, electrical engineering, or a related technical field.

Dolby Hiring Entity:
Deutschherrnstrasse 15-19 90429
Nuremberg
Germany

#LI-MS2