Speechmatics Releases Breakthrough Autonomous Speech Recognition Software Able to Understand All Voices

Speechmatics, a leading British company in deep learning and speech recognition technology, has launched its ‘Autonomous Speech Recognition’ software. Using the latest techniques in deep learning and with the introduction of new breakthrough self-supervised models, Speechmatics claims to outperform Amazon, Apple, Google, and Microsoft in accuracy and ability to understand all voices.

Boston-based ReadSpeaker, an independent voice developer, unveiled its Neural TTS Engine, a Deep Neural Network (DNN) engine powering the company’s voice enablement platform. With the Neural TTS Engine, ReadSpeaker can generate superior-quality speech using DNN-trained models, and produce voice solutions at a fraction of the time required by legacy offerings. The enhanced platform comes amid explosive interest among businesses to use voice technology to best engage audiences in an increasingly contactless world.

Following the announcement in February 2018 of the launch of Global English, a single English language pack supporting all major English accents for use in speech-to-text transcription, speech recognition technology specialists Speechmatics have now announced its Next Generation languages update, which the company says brings an improved accuracy of up to 16% across many of its core languages, thanks to the use of new machine learning algorithms.