RecoMadeEasy ® AudioVisual Engine Demo 2018
RecoMadeEasy ® Speech Recognition and Speaker Recognition demonstration. This video shows our full speech diarization engine which is a part of the RecoMadeEasy ® AudioVisual engine of Recognition Technologies, Inc. (http://www.recotechnologies.com), capable of doing speech recognition, speaker recognition, speaker segmentation, emotion detection, and face recognition in a single engine with one simple API, all on embedded systems without any need for a server or cloud services. The speech transcription can handle 240,000 unique words in any of English, Mandarin Chines, Arabic, and German languages, with bilingual capabilities, totalling 340,000 words, 240,000 in the main language and an extra 100,000 words in English, capable of full code switching applications. The speaker recognition includes very large scale text and language independent speaker identification, very accurate speaker verification, and speaker segmentation. The engine also provides large scale face recognition and the best published results on emotion detection using speech.
For more information see http://www.RecoTechnologies.com