| 摘要 |
An improved method for estimating the status of human physiological and/or psychological conditions by observing the values of the vocal tract parameters output from a digital speech encoder. The user (40) speaks to his device (41), which transforms the input speech from analog to digital form, performs speech encoding (42) on the derived digital signal, and provides values of speech coding parameters (43) locally for further analysis. The stored mathematical relation, e.g. the user-specific vocal tract transformation matrix (47) is retrieved from the memory and utilized in the calculation of corresponding condition parameters (45). Based on these calculated parameters, an estimation (48) of the present status of user's condition can be derived (48). |
