GreenKey Technologies, creator of an artificial intelligence (AI) and voice-driven collaboration tool for financial market participants, announced today that the U.S. Patent and Trademark Office has accepted its application for improved speed and transcription accuracy of automatic speech transcription based on multiple speech-to-text engines.
The speech recognition engine, GreenKey Scribe, can reduce transcription error rates by 50 percent using the approach documented in the new patent. Scribe uses an array of transcription engines trained on different content domains and automatically selects the engine that will perform best on a particular segment of a conversation. The transcription engines that Scribe leverages can be a combination of models built by GreenKey, engines built by partner transcription providers and cloud transcription services.
Other multi-engine approaches send audio to many engines in parallel and compare the transcriptions. This process can be slow and computationally expensive. Through the method detailed in GreenKey’s new patent, termed “Scribe SwitchBoard,” Scribe can analyze an audio file in less than a second to determine which engine will transcribe it best, without having to send it to all of them.
Tom Schady, GreenKey Chief Technology Officer, said: “We’re in an exciting era of rapid advancement along several avenues of speech AI. In order to deliver the best, most accurate results for our customers, we complement our technology with the most cutting-edge solutions available today. This allows us to embrace innovation across the entire spectrum of automatic speech recognition (ASR).”
GreenKey Scribe was designed to help firms transcribe noisy, jargon-heavy audio with high accuracy in a variety of industries, from Finance to Emergency Services, where common cloud transcription solutions typically perform poorly due to the audio content. With its latest transcription method, Scribe enables firms to maintain high transcription accuracy across a variety of accents, languages and audio quality environments.
Dr. Tejas Shastry, Chief Data Scientist for GreenKey Technologies, said: “Scribe’s SwitchBoard changes the paradigm on how we train AI to transcribe human speech. Firms can train their own customized models, and Scribe can leverage them all together. Beyond increasing accuracy, it opens up the possibility to transcribe multi-language conversations with a single engine.”