Wikipedia managers have joined the KTH Royal Institute of Technology in Sweden to develop which will be the first platform of voice synthesis prepared with the collaboration of all users that will allow you to not only read Wikipedia, but that “talk” to the Wikipedia.
The idea of its creators is to facilitate the use of their vast content to people who have difficulty reading or who have visual disabilities. Although the platform is optimized for Wikipedia It will be an Open Source development anyone can take advantage.
First in Swedish and English
Joakim Gustafson, Professor of synthesis technologies – do not confuse with voice-recognition – in the KTH, he explained that will first focus on the Swedish, since it is the language that have more resources to be able to start work. Starting from there they will begin to work immediately in English, as they will be relatively easy to implement due to the large number of available language resources.
In the project Wikispeech other associations collaborate and companies such as the Swedish Post and Telecom L’autorithy (PTS) or the Swedish firm also specialized in this segment STTS. According to the PTS, 25% of all Wikipedia users “need or prefer the text in spoken form” in the form they have to read.
Dictate that information directly and optimized is therefore the objective of this project. The idea is to provide support for the Swedish, English and Arabic -at least in a rudimentary way – this year then go support gradually no less than 280 languages in which there are articles in Wikipedia.