A locally developed text-to-speech software known as Ariana 3 was recently launched by an Iranian tech company.
Text to speech, abbreviated as TTS, is a form of speech synthesis that converts text into spoken voice output. Text to speech systems were first developed to aid the visually impaired by offering a computer-generated spoken voice that would "read" text to the user.
Unveiled last month, the latest version of the specialized software was released by a science-based firm known as Asre Gooyesh Pardaz, IRNA reported.
Hadi Veysi, the software project manager, said although several TTS software versions in English are of high quality, the previous versions of the same software developed in Persian had one major drawback and that was the digitized robotic voice.
He explained that the first versions of the software worked with JAWS screen reader for targeted users with poor eyesight.
JAWS screen reader is the world's most popular screen reader developed for computer users whose vision loss prevents them from seeing the screen.
"The current version has been developed and modified over six years. The most outstanding factor of the latest version is that the sound is similar to natural speech and easily understandable," he said.
Veysi explained that for the software to read a text out loud, it makes use of mathematical modelling and texts can be read in the voice of a man or a woman.
"The recordings were performed by professional voice actors. In any given context, the new and advanced version has the ability to tell apart the different meaning of homographs and pronounce them accordingly. It can also detect and read numbers in prices, dates and the time," he said.
According to the developer, Ariana 3 can be particularly useful for creating automated telephone messages.
Veysi added that the software will be particularly useful for the visually impaired and the elderly. News outlets could add the feature to their websites to allow for easy reading for the groups mentioned.
It can also be of use in smartphone applications that read audiobooks and text messages.
The new software was tested on the company's website www.farsireader.com and it did not fail to deliver. However, there were some errors in the pronunciations of certain words and intonations that would probably be corrected in later updates.
Asre Gooyesh Pardaz is the first Iranian company specializing in artificial intelligence and voice signals. With a team of specialists from Sharif University of Technology, the company started work in 2003 and has since released several products and services in the area of voice detection and processing specifically in the Persian language.
Some notable products released before are a voice recognition and dictation software known as Nevisa, a software for auto-suggestion and auto-correction in typing as well as a Persian to Arabic and Persian to English interpreting software.
