PhD Position: Diffusion-based Deep Generative Models for Speech Signal Processing (m/f/d)

Duties include academic services in the project named above. Research associates may also pursue independent research and further academic qualifications.
The primary focus of the Signal Processing (SP) research group is to develop innovative machine learning methods for speech signal processing. Our applications range across speech communication devices including but not limited to teleconferencing systems, assistive listening devices, and voice-controlled assistants. In this specific project, we will explore the use of score-based diffusion models to enhance speech using the capabilities of generative AI. Typical responsibilities for a Research Associate include authoring scientific publications and traveling to conferences and workshops to present the findings. We are seeking a highly motivated individual who is passionate about engaging in cutting-edge research within a supportive and pleasant working atmosphere.

A university degree in a relevant field.
Examples of a university degree in a relevant field include Computer Science, Data Science, and Electrical Engineering. Excellent knowledge in statistical signal processing and machine learning is required, along with good programming skills in Python, experience with modern machine learning libraries, and experience with speech processing and audio processing. Fluent English, both spoken and written, and good communication skills are mandatory. Knowledge of German is helpful; for non-native German speakers, we expect the willingness to learn German.
Interested candidates must have obtained a Master's degree prior to the commencement of employment. Interested candidates are requested to attach a copy of their degree certificates and a transcript of records to their online application.

The Signal Processing research group at the Universität Hamburg has a 3-year opening funded by the German Research Foundation (DFG) on diffusion-based speech processing.