Case Study
Timehouse gives the Finnish Federation of the Visually Impaired our Mikropuhe speech synthetiser after 30+ years of collaboration
- C++ programming
- Machine-learning led parsing of phrases
- First Finnish-language speech synthesiser
Timehouse and the Finnish Finnish Federation of the Visually Impaired have worked together since 1989, when the very first version of the Finnish language speech synthetiser, MacinPuhe came out for Apple Macintosh-computers. Mikropuhe for PC-computers came year later. Now Mikropuhe has been updgraded to be compatible with 64-bit computing and given to the Federation for distribution.
Company & collaboration
The Finnish Federation of the Visually Impaired is a non-profit umbrella organisation that brings together all Finnish associations that advocate for the visually impaired. Experts in all matters visual impairments and sight, the Federation advocates for all blind and visually impaired people, and can trace its roots all the way back to 1928.
How was Timehouse selected as a partner?
Timehouse has worked with the Federation since the early days of us developing Mikropuhe (the first Finnish language speech synthetiser that made use of the computer sound card) in the 1980s. The Federation has recommended Mikropuhe to its members as Finnish-language speech synthesiser and has worked with Timehouse on adapting Mikropuhe for visually impaired users.
Challenge
Mikropuhe has long been the speech synthetiser the Federation recommends for screen reading in Finnish. But to ensure Mikropuhe kept working properly, the system needed technical upgrades to make it compatible with the latest machines and screen readers. The Federation also wanted to directly distribute Mikropuhe to its members and integrate the software into its other solutions.
Solution
The reason that Mikropuhe was recommended by the Federation in the first place was its ability to parse Finnish language when spoken at a rapid pace. Rapid speech is how visually impaired people communicate when using Mikropuhe to navigate their computers and devices. In the 1990s and 2000s when other Finnish speech synthetisisers first came on the market, they were not as effective at analysing fast speech. Mikropuhe is uniquely capable in parsing through Finnish language in an accurate, non-speculative way due to the way it has been coded. In Mikropuhe, text-to-speech is integrated to its core and it is able to accurately synthesise the very fast speech that visually impaired users use with minimal errors.
“The machine voice that goes with the screen reader must work quickly and flawlessly. Many users of screen readers prefer using Mikropuhe, especially in tasks that require correct pronunciation of every word and character. Thanks to the project’s financial backers and Timehouse Oy, we will be able to distribute the up-to-date version of Mikropuhee without license fees in the future.”
Iiro Nummela
Service manager
Federation of the Visually Impaired
Also, as Timehouse and the Federation had closely collaborated on Mikropuhe over the decades, the software was optimised for the needs of the Federation and its stakeholders. Timehouse used real-time live user feedback to improve Mikropuhe as an accessibility solution.
This latest collaboration with the Federation included the publishing of a brand new 64-bit optimised version of Mikropuhe that would make it compatible with the latest screen reading software in Windows and Linux operating systems. With this new version, the Federation also has the ability to use Mikropuhe in its own accessibility software and products.
The Mikropuhe project was joint funded by the Federation, The Library Association for the Visually Impaired (Näkövammaisten kirjastoyhdistys ry) and the Tampere region’s Support Association for the Visually Impaired (Näkövammaisten tukisäätiö).
In a nutshell
- C++ coding
- Parsing language accurately through machine learning models
- First commercial Finnish language speech synthesiser ever developed
Upgrading Mikropuhe for latest operating systems
64-bit computing upgrade
30+ year accessibility collaboration
