Thesis topic proposal
Multilingual and cross-lingual speech recognition using Deep Neural Networks


Institute: University of Szeged
computer sciences
PhD School in Computer Science

Thesis supervisor: László Tóth
Location of studies (in Hungarian): SZTE
Abbreviation of location of studies: SZTE

Description of the research topic:

The current speech recognition technologies are language independent. That is, we apply the same technology for all languages. However, as the languages are different, we use a separate training database for each language. The topic of multilingual and cross-lingual speech recognition deals with the question of whether using data from other languages can help the recognition of a given language. On one hand, increasing the
amount of training data is always good. On the other hand, as the
additional data is from another language, it may also confuse the
recognizer, so the optimal way of making use of the samples coming from the other language(s) is not trivial at all.

Required language skills: english
Number of students who can be accepted: 1

Deadline for application: 2019-03-15

All rights reserved © 2007, Hungarian Doctoral Council. Doctoral Council registration number at commissioner for data protection: 02003/0001. Program version: 2.2358 ( 2017. X. 31. )