| 摘要 |
To date, I am not aware of any method which utilizes available data schemes to directly produce a grammar for configuring a piece of voice recognition software so as thereby to improve the level of recognition and to directly produce instances of the data scheme. The level of recognition for the spoken input is very high and will in practice result in this form of detection being implemented for geodata (compulsion for in-situ detection). Areas of application may be street tree documentation or street state detection, for example. For document review, areas of application in the field of inspection documentation - for example for the maintenance and inspection of complex technical facilities - are probable.
|
