Multilingual Speech Recognition with Semantic Accuracy
Multi-lingual Speech Recognition and Theme-Semanteme Analysis Method and Device
Tags: National Cheng Kung University, Taiwan, Computing Technology, Consumer Goods & Services
This technology focuses on multilingual speech recognition combined with subject semantic analysis. It uses a speech recognizer to convert voice input into a pinyin string, mapping it to multiple languages, and a semantic analyzer to refine the sentence through correction processes. This system addresses the challenges of multilingual dialogue and mispronunciation, enhancing sentence accuracy. Applications include voice dialing, navigation, and device control across diverse linguistic contexts. It aims to improve cross-lingual communication through more accurate speech-to-text conversion.
IP Type or Form Factor: Patent Granted; Process & Method; Software & Algorithm
TRL: 5 - prototype ready for testing in intended environment
Industry or Tech Area: Big Data Analytics & Simulations; Home Appliances & Consumer Electronics
