|
SBE 2020: Submission Detail
| ID Number: |
55 |
| Title: |
Real-World Speech Recognition |
| Lead Author: |
Rubin, Philip |
| Abstract: |
Speech recognition would seem, to many, to be a scientific/technical problem that has been solved. Inexpensive recognition systems are commonly available for personal computers and mobile devices. Why then is the use of such a potentially enabling technology not as ubiquitous as past predictions would have led us to believe? Note that I have typed this into my computer, not spoken to it. One rarely sees people talking to their computers, unless they are skyping, although the recognition ( talking typewriter ) technology supposedly has been mastered. Unfortunately, recognition performance is severely limited by real-world constraints. Ambient noise, variability in the clarity of a speakers voice due to age, speaker style, infirmity, and a host of other conditions limit the practical and reliable use of speech interfaces. Speech is more informal and capricious than algorithmic approaches are designed to handle. In addition, we help disambiguate such ephemeral information by using as many contextual, communicative cues as are available to us, including facial information, gesture, indications of emotion, and situational indicators. The challenge is to mount a sustained, focused effort to develop recognition systems (speech, gesture, facial information, emotion, semantic, etc.) that work reliably in real-world conditions, from the workplace to the battlefield. |
| PDF: |
Rubin_Philip_55.pdf |
SBE 2020 Home
|
