The goal of SmarTerp (Seamless Management and Automation of Resources and Tools for an Efficient Remote Simultaneous Interpreting) is to reduce inefficiencies in interpreting by developing a set of AI-powered tools embedded in a Remote Simultaneous Interpreting system that automates the human task of extracting information in real-time to prevent the mistakes and loss of quality derived from the adoption of remote technologies.
The project will leverage on the following technologies:
- Licode (Open Source WebRTC-based MCU) to complete the architecture of the RSI system prototype with real-time communications.
- Automatic Speech Recognition of audio streams based on Open Source code.
- Knowledge-graph based Natural Language Processing architecture for the recognition of named entities, time expressions and relationships between terms from texts in different natural languages, to work in real-time conditions.
TThe role of SpeechTek is to improve the performance of its speech recognition engine for the languages covered by the project, both in terms of accuracy and real time processing, and its integration into a coherent and usable AI-powered platform
Partners:
- Eidolon SL
- FBK
- UNIVERSIDAD POLITÉCNICA DE MADRID
- University of Bologna
Dates: January 2021 – December 2021
