20th International Conference on Speech and Computer


Big Data in Speech Computation


Big Data is an emerging topic in a broad variety of research areas.

Especially deep learning methods gained huge successes in a broad area of applications such as speech recognition, social signal analyses and natural language processing. Big data furthermore allows to conduct predictive analytics, user behaviour analytics to enhance the system knowledge on user groups.Along with the growing data and new methods also the problem of data capturing, data storage and data analysis. Many of these issues demands a manual annotation, which is not feasible in the case of big data. Thus, methods to allow a proper (semi-)automatic annotation of big data are needed.

On the other hand for emotional analyses and interaction analyses of speech data the capturing of big data itself is challenging. Methods to generate speech data from learned representations may help in this context.

In this special session we call for for research papers dealing and discussing the above mentioned issues. The topics include (but are not limited to): deep learning, transfer learning, (semi-)automatic speech data analyses (for big data), language modelling and big data generative adversarial networks.


Anna Esposito (Università della Campnania "Luigi Vanvitelli" , Italy) - iiass.annaesp@tin.it

Friedhelm Schwenker (University Ulm, Germany) - friedhelm.schwenker@uni-ulm.de

Ingo Siegert (Otto von Guericke University Magdeburg, Germany) - ingo.siegert@ovgu.de