what3words Voice API features Speechmatics automatic speech recognition technology

what3words has entered a partnership with automated speech recognition software provider Speechmatics for users to enter their what3words address by voice in their own language, the company announced.

Under the partnership, the what3words Voice API leverages machine learning and automatic speech recognition from Speechmatics to deliver an API service that integrates automatic speech recognition technology with in-built what3words post-processing to interact with voice-controlled devices.

The two presented the end-to-end speech recognition and post-processing API at CES 2020 in Las Vegas.

The voice API lets the user speak three words into an application or service and returns the address and GPS coordinates in a single API call. By embedding Speechmatics’ technology, more users around the globe will be able to use voice devices to communicate precise locations in their own language.

what3words broke down the world into a grid of 3m squares where each square has a unique address made of three words which, the company says, is optimized for voice input to deliver a better experience for users that enter traditional street addresses in a voice interface. Often addresses use non-dictionary words or pronunciation that is hard to guess. what3words addresses are unique and placed far apart if similar, which the company claims makes it easier to detect and correct errors based on the what3words AutoSuggest post-processing function.

The API makes it easy to enable what3words-only voice search because it removes the commercial and technical complexity of merging multiple APIs. The product is hosted on servers in multiple international hubs, making the audio stream’s latency low enough to respond to in real time.

“The what3words Voice API in partnership with Speechmatics means that what3words address searches by voice can now all be done in this single product,” says Chris Sheldrick, CEO and Co-founder of what3words. “This makes it the ideal solution for apps and services that want to enable simple address entry via voice, without the speech recognition ambiguity associated with address search through traditional ASR systems – duplicate names, for example.”

“Speechmatics enables its partners to innovate with voice by providing accurate speech recognition regardless of your accent and can learn new languages on demand,” added John Milliken, CEO of Speechmatics. “We’re delighted to support what3words in the launch of our combined API. Using automatic speech recognition to unlock the value in voice data is becoming more widespread and we are excited about the potential for our technology to help what3words create a new global standard for addresses.”

The what3words Voice API powered by Speechmatics currently supports English, Mandarin, Hindi, Japanese, Korean, German, Arabic and Spanish.

Speechmatics secured £6.35 million (USD$8.254 million) in Series A funding in October led by AlbionVC and other angel investors including IQ Capital. The company said it will use the funding to grow global presence in multiple markets.

The automatic speech recognition market is valued at $7.5bn and forecast to reach $21.5bn by 2024.

Speechmatics’ speech recognition technology is available in 29 languages, in the cloud or on-premises. The company recently received a Queen’s Award for Enterprise in the ‘Innovation’ category for 2019.

Automatic speech recognition technology is an important component in human-machine interaction and for businesses integrating machine learning and AI into their strategies. Deployments include almost instant transcription of audio files, live subtitling in broadcast, and the conversion of call center recordings into text.

Speechmatics recently appointed John Milliken as CEO to strategize for global expansion. The funding received will be invested in product development and geographical expansion that includes opening new offices in Denver USA, Brno in Czech Republic, and Chennai in India.

“We’ve invested heavily into our product and machine learning capabilities to provide our customers with what we believe is the best speech recognition technology on the market,” he said.

“We recognize that businesses are increasingly aware of the importance of unlocking their voice data and utilizing it for core business functions such as compliance, analytics, training and improving customer experience. This latest investment opens up not only global expansion but the opportunity for new product development and traction in new markets.”

“Speech software is a large and rapidly growing market and we believe Speechmatics’ small footprint language models, capable of flexible deployment in the cloud, on-premises or on-device will become of strategic importance in the evolving landscape. We believe Speechmatics, led by John, is poised to become a significant global software company,” said Robert Whitby-Smith, Partner at AlbionVC.

Article Topics

API  |  biometrics  |  machine learning  |  speech recognition  |  Speechmatics  |  voice biometrics