High Accuracy Speaker-Independent Multilingual Automatic Speech Recognition System

The goal of the project is to develop a speaker-independent multilingual template-based automatic speech recognition (ASR) system, which flexibly scales to new languages and speakers, and performs robustly for multilingual vocabularies and non-native speakers in noisy environments. This shall allow the company to quickly deploy its applications for the Hospitality&Food service industry to any linguistic area, and thus open new perspectives on the international market.
Perceptive and Cognitive Systems
Idiap Research Institute
Innosuisse - Swiss Innovation Agency
Nov 01, 2013
Mar 31, 2018