I later improved by pulling apart a gyroscopic mouse sticking the gyroscopes on a cap. I used to use a head pointer device called trackIR ( it used infrared ) and that was really decent. Second thing to bear in mind is that programmers use the mouse a lot. I christened the project 'spascii' (ala 'speak ASCII') but I never started coding it. Maybe one day I will have a crack at it if no one else does. You could conceivably rack up a huge speed dictating this way, very possibly faster than a trained typist, bearing in mind you use all those crazy symbols when you code. Then you speak to your computer using a sequence of nonsense 'la-lu-be-ni-pa-ta-bu-rru.' and it types as you go This will leave a lot of slots left over which you can customise. You then adapt Sphinx using ONLY these 10+10 phonemes, and 100 word dictionaryĪnd map a different keyboard character to each slot. So, after weeding out ones which are too close to their neighbour, you are left with at least 100 distinct phoneme pairs. English uses probably close to 20 consonant sounds and 13 vowel sounds, though some of these are dipthong/tripthong. going horizontally you have the consonant phonemes. I got as far as creating a blueprint for a working solution, which goes as follows: Its called Enhanced Dictation if you download some data so the recognition takes. I doubt there is a solution that I haven't tried. MacSpeech Dictate was the predecessor to Dragon for Mac and Dragon Dictate. I have been plagued with RSI for 10 years.
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |