PhD position "Phase Retrieval in Imaging and Speech Enhancement"

The position is a collaboration between Signal Processing (SP) at the Department of Informatics of the Universität Hamburg, Germany and the German Electron Synchroton DESY in Hamburg.

Prof. Dr.-Ing. Timo Gerkmann:
Prof. Dr. Henry Chapman:

Data Science is the science of extracting knowledge from typically large amounts of complex data. The increasing level of automation and the increasing number and resolution of sensors in scientific large-scale experiments result in large, heterogeneous and highly complex data collections. Therefore, Data Science is a key technology in modern natural sciences. Data-intensive research in experimental and theoretical science at the Science City Bahrenfeld, Hamburg, Germany requires beyond off-the-shelf software solutions for data management, processing and analysis. Tailored or even completely new computational methods are needed and we will explore new innovative ideas in computer science and applied mathematics. To develop these methods, scientists have to be highly skilled in their corresponding natural-science background and computer science alike. The Data Science in Hamburg Helmholtz Graduate School DASHH has been established to address this highly interdisciplinary need. DASHH bundles competences of scientists with high international reputation in basic research for the structure of matter, computer science and applied mathematics in Hamburg in a new and unique fashion.

Activities and responsibilities

Advanced sources such as free-electron lasers produce intense and coherent beams of X-rays that are opening up new possibilities to image biological materials, such as single molecules, at atomic resolution. Since atomic-resolution lenses do not exist such methods usually rely upon retrieving the structural information encoded in the far-field coherent diffraction pattern [Shechtman et al, 2015]. This intensity pattern corresponds to the Fourier magnitude of the object, and is thus an incomplete measurement since the spectral phase cannot be measured. To reconstruct the original structure, the missing phase information needs to be retrieved. A very similar problem of phase retrieval is encountered in speech source separation and enhancement, which has received increasing attention recently [Gerkmann et al, 2015]. Just in the last few years, modern machine learning methods have been successfully applied to the problem aiming to obtain faster and more accurate results.

The goal of this project is to explore common concepts in X-ray imaging and speech enhancement. Recent advances in the application of machine learning in speech analysis will be transferred and tailored to X-ray imaging, obtaining deeper insights into this challenging inverse problem and providing new opportunities for high-resolution structure determination of biological systems.

Qualification profile

In order to apply you need to have a Master's Degree (or an estimated date of Graduation within the same year of application) in computer science, applied mathematics or natural sciences, preferably with an interdisciplinary training at the interface of natural and computer science or mathematics. Degrees from foreign universities and master degrees in non-research oriented study programs (Fachhochschule) might be eligible. For further information, consult the doctorate regulations (Promotionsordnung) of the respective partner university.
All prospective students must submit proof of adequate English proficiency. The following proofs are accepted:
  • A university degree in an academic study program taught in English
  • TOEFL paper and pencil test: 580 points with a score of 45 in the essay (TWE score)
  • TOEFL internet-based test: 92 points with at least 22 points per skill (listening, writing, speaking, reading)
  • Cambridge Certificate of Advanced English (CAE): Results A, B, C
  • Proof of at least five years of English lessons at a German/Austrian/Swiss school
  • International English Language Testing System (IELTS) – Academics: Band 6,5, Good Competent User, with at least 6,0 per skill
  • Proof of English proficiency with the Level B2 (Common European Framework of Reference for Languages) from a recognized institution
TOEFL ITP and TOEIC are not accepted. Acceptance of other than the above-mentioned language certificates remains subject to approval by the admissions board.

We offer

The graduate school covers data challenges from application fields such as structural biology, particle physics, material science and ultrafast X-ray science. These data challenges have in common that they cannot be addressed with standard computational methods, but require modern data- and information-science techniques instead. Several areas from computer science and mathematics play important roles: data management and engineering, machine learning and data analytics, signal and image processing, algorithm design, optimization and simulation, software engineering and automation and control systems.

Send application to

While applying for the job please refer to jobvector

About Universität Hamburg

The focus of the Signal Processing group is on algorithms for processing speech and audio signals with applications in hearing aids, communication devices, and human-machine interfaces.
More about Universität Hamburg