1. Text-to-Speech, Speech Synthesis
2. Speech Recognition
3. Language Technologies
4. Voice Services
5. Voice Portals, IVR
6. Accessibility
7. FAX

Text-to-Speech, Speech Synthesis

Speech Synthesis is the artificial production of human speech (Speech Synthesis). It is a functionality that many application fields have interested in, as it gives the capability of reading unlimited and dynamic messages without any human interaction. Synthetic speech can be produced by concatenating small or large voice samples or by modeling the procedure of human articulation and vocal tract. Based on the approach, voice qualities that can be achieved vary from extremely natural to robotic hues. However, almost all are capable to generate understandable speech by transmitting the text content to the audience via the speech signal. It is a matter of taste and application the choice of the right and most appropriate voice.

e-rhetor has long experience in speech synthesis systems. The outcome of its research developments is the PhonAesthesia Language Platform, which supports among other technologies, the creation of all types of synthetic voices. In close collaboration with international voice houses (Loquendo, Acapela), e-rhetor offers a plethora of voice and languages to select from. See:

• pesto! TSS, the free, small-footprint and flexible Greek synthesizer
• PhonAesthesia Visual for screen reading applications
• the on-line text-to-speech service WebRhetor
• the on-line voice RSS alert service WebRhetor Feed: VoiceRSS
• the free telephony RSS voice browser WebRhetor Feed: PhoneRSS that lets you voice browsing all popular Greek sites
• Voice Factory, a service for creating natural voices in small time
• Loquendo TTS, enhanced for Greek by PhonAesthesia Language Platform

Speech Recognition

Speech Recognition is the process of converting spoken text to words (Speech Recognition). This "understanding" procedure depends on many factors and usually it is computer driven by limited language knowledge (grammars).

e-rhetor has developed speech recognition language models and has great experience in Nuance Voice Platform and Loquendo Speech Suite, which are being used in its services. See:

• Loquendo ASR
• WebRhetor Feed: PhoneRSS free telephony RSS voice browsing service
• the interactive voice response system IVRhetor

Language Technologies

e-rhetor PhonAesthesia Language Platform is a system with language knowledge for processing text and voice data. Some application fields that is being used are:

• Speech Synthesis
• Speech Recognition
• Grapheme to phoneme conversion
• Speech segmentation
• Stress prediction
• Greeklish to Greek conversion
• Text normalization (numerics, abbreviations, acronyns) with grammatical declensions

Voice Services

See:

• WebRhetor
• WebRhetor Feed: VoiceRSS
• WebRhetor Feed: PhoneRSS

Voice Portals, IVR

See:

• WebRhetor Feed: PhoneRSS
• BLIND.GR
• IVRhetor

Accessibility

See:

• PhonAesthesia Visual
• BLIND.GR

FAX

See:

• FaxRhetor