Improving speech playback using time-compression and speech recognition
Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
Methods for the visually impaired to retrieve audio information efficiently
ICCHP'10 Proceedings of the 12th international conference on Computers helping people with special needs
ICCHP'10 Proceedings of the 12th international conference on Computers helping people with special needs: Part I
| Hi-index | 0.00 |
We propose an “oblique” listening system in the English language for the blind and visually impaired by controlling the speed of Text-to-Speech based on parts of speech in which the important parts are synthesized 'relatively' slower while the unimportant at the maximum speed. In the system evaluation experiment, English natives with vision impairments were required to listen to three short passages of Text-to-Speech at the three types of speed and answer the questions of comprehension after the base-line speed was calculated from the measured maximum speed of recognizing 'a word' and that of recognizing 'a sentence' for each subject. We controlled the Text-to-Speech speed: base-line, 'oblique listening' speed, and simple high speed, by which the duration is equal to the one by our system but without any speed variation. The results show that the “oblique” listening system is better than a simple high speed system even though not exceeding the base-line