Speech Technologies

Sakhr provides software for Arabic Text-to-Speech (TTS) and Automatic Speech Recognition (ASR). Sakhr TTS converts Arabic text into a natural, human-sounding synthetic voice. Sakhr ASR engine recognizes spoken Arabic in different accents, without requiring user training. Sakhr ASR & TTS solutions include:

• Speech to Speech Translation
• Mobile Applications
• Call Center IVR (Interactive Voice Response)
• Directory Assistance
• Arabic Language Learning

Text-to-Speech (TTS)

The Sakhr TTS engine converts Arabic text into human voice. Sakhr TTS is the industry leader in synthesizing a natural, human-sounding Arabic voice. Sakhr TTS leverages the company’s 28+ years of research and corpora development in Arabic natural language processing (NLP). This research is critical to overcoming the TTS challenges of the Arabic language, such as lack of diacritics and punctuation. Features:

• Most natural, intelligible Arabic voice in market
• Embedded TTS engine for mobile devices
• Rules-based powerful diacritizer with 97% accuracy
• Normalizes ambiguous text such as dates, time, currencies, abbreviations
• Unlimited vocabulary, text size, raw text, phonetic and prosodic input

Speech Recognition (ASR)

Sakhr’s ASR engine is the core technology recognizing spoken Arabic across a wide range of utterances, noisy environments, and Arabic accents from the Arabian Gulf, Egypt and Levant. This engine is industry-standards compliant and can be integrated with mobile client or telephony applications. Features:

• Speaker independent
• High accuracy recognition for large or small vocabularies
• Supports different Arabic accents such as Gulf, Levant & Egyptian
• Integrates with telephony systems for IVR, directory assistance, voice dialing
• Integrates with mobile devices & client applications
• Supports Microsoft Speech Server
• Supports proprietary API and MS SAPI 5.1
• Supports open standards such as MS SALT