Peabody, MA — (press release excerpt) — ScanSoft, Inc. today announced the RealSpeak Compact SDK, a text-to-speech (TTS) solution for automotive and mobile applications. The new SDK allows developers to add human-sounding speech capabilities to applications developed for devices with small processing and memory capabilities, including those based upon Microsoft Windows CE and Linux. The system supports US English, French, German, Italian, Japanese, Castilian Spanish, Mexican Spanish, Brazilian Portuguese, and Swedish languages, running on NEC MIPS, Hitachi SH3, SH4 and Intel x86, StrongARM, and XScale chipsets, depending upon SDK configuration.

Until now, the limited resources of small-footprint devices relegated applications to formant technology for speech, which uses machine-generated sound to create words — utilizing less processing resources but resulting in machine-like speech. RealSpeak Compact is significant because it delivers to small-footprint devices the TTS quality previously reserved for high-end server and telephony applications. RealSpeak Compact is based upon sophisticated concatenative technology, which combines segments of real human speech to form new words and phrases. This approach is already the standard for phone-based unified messaging, interactive voice response and voice-portal applications.

The RealSpeak Compact SDK extends ScanSoft's established leadership in TTS technology, joining the RealSpeak Telephony SDK and RealSpeak PC SDK. RealSpeak Telephony holds a more than 44% market share in the North American market for speech-enabled unified messaging, IVR and voice-portal applications, and is the voice for the official information portals of the 2002 FIFA World Cup and the 24 Hours of Le Mans. The RealSpeak PC SDK is the premier solution for speech-enabling any Microsoft Windows application, and is used in a range of office, game, accessibility and educational titles.

RealSpeak Compact delivers high-quality speech with smaller disk and memory requirements than competing solutions. The system is ideal for adding speech to embedded device applications, bringing to mobile applications the same high quality sound, inflection, tone and pace previously available only on server-based systems. Text-to-speech is particularly important for small- and no-display devices, providing an effective way to deliver interactive prompts as well Web, email and wireless content.

 
This article was originally published on LinuxDevices.com and has been donated to the open source community by QuinStreet Inc. Please visit LinuxToday.com for up-to-date news and articles about Linux and open source.