The flat-toned robotic American sounding voice, used in the ACAT (Assistive Context-Aware Toolkit) system, developed by Intel and used by Stephen Hawking, has become instantly recognisable as the voice of one of the world's leading scientific brains. Now, however, it is being made available to download for any and all to experiment with.
The toolkit is being made available on an open-source basis and is available to download. It is designed to be used in conjunction with a webcam, where the 'speaker' issues commands by twitching their cheek, just like Professor Hawking has to do before uttering so much as "hello."
According to Wired, it took Hawking as much as 20-minutes just to write a three-word greeting, three years ago. Now, however, it uses SwiftKey to auto-correct the text based on his research papers, so if he types "black" the software will automatically assume that the next word will be "hole". The idea behind releasing the toolkit in this way is to encourage developers to take a look at its functionality and hopefully continue to improve the speed at which it can translate.
We expect it might take folks some time to master the subtle cheek-twitch commands. We look forward to seeing what improvements can be made. Stay tuned to Buro 24/7 Middle East for the latest as it happens...