SpeechWorks Adds New Voices to Speechify

BOSTON, MA - SpeechWorks International Inc. (Nasdaq: SPWX), a developer of speech technologies and services, introduced three new languages for its Speechify text-to-speech (TTS) software. Organizations are now able to deliver text-based information in an audio format in Australian English, Canadian French and Americas Spanish. Speechify 2.1 software now supports ten languages and dialects with voices in U.S. English, Continental French, Latin American Spanish, UK English, German, Japanese, Brazilian Portuguese, Americas Spanish, Australian English and Canadian French. "Microsoft has selected Speechify as the TTS engine for the Microsoft Speech platform due to its ability to support high quality, natural sounding synthetic speech in multiple languages," said James Mastan, Director of Marketing, Microsoft Speech Technologies Group at Microsoft. "With the Speechify TTS engine integrated into Microsoft's SALT-based Speech Platform, organizations will be able to reach a broader audience of mobile users for both telephony and multimodal applications." The product leverages unit selection algorithms for modeling voice quality, overall rhythm and intonation to produce a synthetic voice. Speechify also applies front-end processing, to predict textual ambiguities in a sentence (e.g. "St." meaning "street" or "saint") or distinguish between words with identical spellings but different pronunciations (e.g. "record" or "read"). This front-end processing now includes an address tag, which applies a special set of processing rules to portions of text that have been identified and tagged in a database. This results in improved rendering of street addresses, which is useful in location and navigation applications. Organizations can repurpose existing data for a mobile customer base, instantly deliver dynamically changing information in an audio format for each caller and decrease costs by reducing the need for pre-recorded audio. Callers experience a smooth sounding voice that delivers spoken information as a human would. "SpeechWorks speech recognition and TTS technologies support our continued efforts to improve customer service in our Australian contact centers," said Hank Jongen, National Manager at Centrelink, the largest single purpose contact center network in the southern hemisphere, processing 24 million telephone calls and distributing AU$51.7 billion in social security payments each year. "Providing our callers with a natural sounding voice to relay information about workers' wages is crucial. Speechify's Australian English voice truly mimics the Australian accent and sounds incredibly natural and smooth. We are excited to implement this new technology and believe that it will allow us to reduce costs in our contact centers while simultaneously providing better service to our callers." "As SpeechWorks introduces new languages for Speechify, we continue to focus on two critical measurements in evaluating TTS performance -- naturalness and intelligibility," said Steve Chambers, chief marketing officer at SpeechWorks. "But it's not just the quality of the TTS, it's what organizations can do with this technology. Speechify can be used to create a custom voice to extend an organizations brand identity. In addition, it can be blended with pre-recorded audio to deliver dynamic information naturally, enabling more effective speech applications while also positively impacting the perception of the organization." All Speechify voices, including Australian English, Canadian French and Americas Spanish, can also be made available for Speechify Solo software, SpeechWorks' embedded TTS engine. Using proprietary memory reduction techniques, Speechify Solo offers voice quality that closely mirrors Speechify 2.1 software but in a footprint appropriate for embedded devices. Speechify Solo can be used to read aloud SMS messages, artist names and song titles for in-vehicle entertainment systems, navigation and driving directions, banking transactions in automated teller machines and name reading for voice activated dialing applications. Through SpeechWorks Virtuoso V.I.P. (Voice Identity Program), companies can also create custom voices using Speechify and Speechify Solo to match their recorded voice, creating a branded voice for TTS applications and combining it with recorded prompts.
SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues