ISCA - International Speech
Communication Association


ISCApad Archive  »  2016  »  ISCApad #217  »  Jobs  »  (2016-01-21) Researcher in machine learning at Reykjavik University, Iceland

ISCApad #217

Sunday, July 10, 2016 by Chris Wellekens

6-7 (2016-01-21) Researcher in machine learning at Reykjavik University, Iceland
  

Machine learning and language technology

Reykjavik University is looking for ambitious candidates to work on development and
implementation of speech recognition and other language technologies. The development of
speech recognizers applies machine learning on big data sets of text and speech
recordings. The machine learning that is used in the projects is mostly implemented and
available in open source software. The work typically includes gathering and preparing
data, setting up and configuring the machine learning procedures, run experiments and
design the interface for users and other software. The positions are for one year with
possibility of extensions.

Specialist in machine learning
The main focus of this job is to apply machine learning on big text and speech datasets
to develop and train automatic speech recognizers. The software that is used in the
project is called Kaldi, which demands a high skill in the use of Linux and associated
tools. The work is to set up and evaluate computational models (finite state transducers,
hidden Markov Models and deep neural networks) on large datasets.

Skills:
? BSc/MSc degree in mathematics, statistics, engineering, computer science or similar.
? Knowledge of computational modeling is preferable (e.g. differential equations, neural
networks, linear models).
? Good knowledge of Linux is necessary.
? Good skills in writing and understanding shell scripts are preferable (bash, awk, sed).
? Good programming skills are necessary (e.g. C++, Java or Python).

Researcher in machine learning
The main focus of this job is to carry out research on speech recognition using deep
neural networks. The theoretical part of the work can either concentrate on parameter and
model optimization with respect to speech recognition performance or on learning setup
and model configuration with the aim to automate training of speech recognizers. The
group is already using open source speech recognition solutions of Kaldi, Tensorflow and
Theano. Some systems are already in operation while others are in the process of being
implemented. Design improvements and adaptation will continue in the coming months and
years, so the research will have a very direct practical impact.

Skills:
? MSc/PhD degree in applied mathematic, statistics, computational engineering or computer
science is preferable.
? Knowledge of mathematical modeling is preferable (e.g. differential equations, neural
networks, linear systems).
? Good skills in applying and analyzing algorithms.
? Good knowledge of Linux is preferable.
? Ability to use shell scripts is preferable (bash, awk, sed).
? Good programming skills are necessary (e.g. C++, Java or Python).

For further information contact Jón Guðnason, (jg@ru.is) Assistant Professor in School of
Science and Engineering.
Applications should be submitted before the end of 15th of April 2016 but strong
applications might be considered earlier.
Please submit your application through links provided at:
http://radningar.hr.is/storf/viewjobonweb.aspx?jobid=2729




Back  Top


 Organisation  Events   Membership   Help 
 > Board  > Interspeech  > Join - renew  > Sitemap
 > Legal documents  > Workshops  > Membership directory  > Contact
 > Logos      > FAQ
       > Privacy policy

© Copyright 2024 - ISCA International Speech Communication Association - All right reserved.

Powered by ISCA