Human language text and video signal parallel processing

Project looking for partner(s)
Institution / Department: 
EDI - Institute of Electronics and Computer Science, Riga, Latvia
Country: 
Latvia
Name: 
Kaspars Sudars
Position title: 
Researcher
E-mail: 
Expertise needed / Role in project: 

We are looking for partners with experience in signal and image processing, neural networks or web page programming.

Partners already involved: 

EDI - Institute of Electronics and Computer Science, Riga, Latvia
University of Latvia, Institute of Mathematics and Computer Science, Riga, Latvia
UPG University of Ploiesti, Romania

Abstract: 

In our project we are going to process video signals and human text parallel according to Grounding Language Processing paradigm. First the tools for multi-modal corpora have to be created for gathering necessary data. The statistical and signal processing methods will be explored including deep neural networks. Potential applications for developed methods could be artificial commentator (similar to "Mooney's RoboCup games with commentaries"), some querying system for navigation in maps etc.
We are looking for partners with experience in signal and image processing, neural networks or web page programming.

Keywords: 
signal and image processing, neural networks, language frames
Topic: 
Human Language Understanding: Grounding Language Learning