Cooperation with Laboratory of Language Technology
The Laboratory of Language Technology conducts research in the fields of speech and natural language processing. One of our most important activities is the creation of speech technology applications. This includes end-user oriented speech recognition applications as well as components that are easy to integrate for industrial partners. Although our focus is on Estonian users, most of the software created in the laboratory is not specific to Estonian. The laboratory is a solid supported of open source software.
OUR EXPERTISE
- Speech-to-text
- Speaker and spoken language identification
- Natural language processing (NLP)
- Collecting speech datasets
WHICH PROBLEMS CAN WE SOLVE
- Various challenging speech processing problems
- Information extraction from speech and text
HOW CAN WE HELP YOU
|
Understanding the possibilities |
Consulting |
Development of models |
|---|---|---|
|
The field of NLP has developed in a rapid pace in the last decade and computers can now do many things that were unimaginable before. Our experts can identify use cases that can be solved with modern deep learning techniques. |
Our experts can help your team with the problems of speech and text data collection, picking the right model architecture, making smart use of existing datasets and models via transfer learning, all this for any language of the world. |
We have a long experience in developing and training accurate and usable models for various speech-to-text tasks as well as other language processing tasks, covering tens of different languages. |
REFERENCES AND PREVIOUS PROJECTS
- General-purpose speech-to-text system for Estonian
- VoxLingua107 – a speech dataset of 107 languages