Speech recognition and natural language processing for digital interviews

In the context of a Swiss NSF grant, we seek a PhD student to work on speech recognition and natural language processing for digital interviews. The project is a collaboration between the universities of Neuchâtel and Lausanne, and Idiap Research Institute.

At a high level, we are interested in how candidates for jobs respond in structured selection interviews; in particular how they are able to tell stories about past work situations. More concretely, we will investigate speech recognition architectures suitable for such interviews, and natural language processing solutions that are able to infer the higher level semantics required by our collaborators in the social sciences.

A-priori, given that it is the higher level semantics that are of interest, we expect to make use of recent lexicon-free approaches to speech recognition. We also expect to draw from the rapidly advancing language modelling field with tools such as BERT and its successors. There will be ample opportunity for research in the component technologies, which are all currently pertinent in the general machine learning landscape. In particular, we expect to make advances in the technological interfaces between component technologies, and in the humanist interfaces between the machine learning practitioners and social scientists.

The ideal Ph.D student should have a master (or equivalent) degree in engineering, computer science, physics or applied mathematics. S/he should have a good background in programming (C/C++, Python, scripting languages); practical engineering skills will be particularly appreciated since the work of our collaborators will be dependent on the output of our part of the project. In order to balance the group, we especially encourage female applicants. However, all applications will be judged on merit.

The student will be registered at EPFL and located at Idiap. However, we expect some time to be spent at our collaborators' locations, especially during data collection and annotation.
The position is available from April 2021, funded for four years. Selection will commence immediately and continue until the post is filled.

Interested candidates are invited to submit a cover letter, a detailed CV, and the names of three references through the Idiap online recruitment system:

Speech recognition and natural language processing for digital interviews