SpeechTek Open positions

DEVELOPMENT OF END-TO-END AUTOMATIC SPEECH RECOGNITION SYSTEMS

A position for fresh graduate is open within the SpeechTek research unit in the field of automatic speech recognition. In particular, the position is preparatory for students interested in pursuing a PhD. A PhD grant on the same topic will be opened at University of Trento in the next months.

Goals:

The goal of the work will consist in porting to different domains and languages an end-to-end speech recognition system based on ESPnet (https://github.com/espnet/espnet). Several approaches will be investigated, mostly addressing the usage of neural network adaptation techniques and of transfer learning.

The work will be coordinated by experts in automatic speech recognition, using both internal resources (data, GPUs, HPC computing) and the cloud.

Required knowledge and skills:

  • Msc degree, preferably in computer science or electronic engineering
  • Knowledge of the linux operating system, linux shell
  • Fluency in Python.
  • Knowledge of the fundamentals of deep learning.

Contract duration: 9-12 months. Init period: February – March 2019.

Gross salary:  1800 – 2100 euro per month

Application: send CV and cover letter to Marco Matassoni (matasso@fbk.eu)

 

INTERNSHIPS – MASTER THESIS

In SpeechTek, we have several research projects for master students interested in internships of 3-4 months on the use of artificial intelligence for

  • automatic speech recognition,
  • language learning and proficiency estimation,
  • speaker diarization and verification,
  • and other speech related fields.

Students will have access to our resources and will work in tight collaboration with our senior researchers, addressing very challenging and advanced research issues.
For further information, please contact Daniele Falavigna, head of SpeechTek.

Back