Phonito™ Processing
Introducing Phonito™ Processing — The Cleanest Breakthrough In Speech Recognition
Whether you’re at home, at work, in a restaurant, or on the road, there is always background noise — a nightmare for speech recognizers. In this noisy world, Red Shift’s proprietary Phonito™ Processing technology hears through it. It separates out the garbage in a new, orthogonal approach to current state-of-the-art speech signal processing techniques. To this end Red Shift made the RASR™ -SuperCharger, which has been shown to improve the accuracy of commercial speech engine when tested in the presence of various levels of noise.
When exploited for speech recognition, Phonito™ Processing provides clearer acoustic features from relatively short frames enabling faster decoding times, higher confidence scores and shorter latency. The result of achieving improved accuracy quicker than ever is manifested in RASR™, a real-time cloud-based speech recognizer.
Steps are under-way to verify that using Phonito™ Processing in speech synthesizers improves their speed and quality.