METHOD FOR EXTRACTING ACOUSTIC FEATURE QUANTITY IN DISEASE INFERENCE PROGRAM, AND DISEASE INFERENCE PROGRAM AND DEVICE USING SAID ACOUSTIC FEATURE QUANTITY

摘要

The present invention provides: an inference device that is for inferring multiple mental and neuronal diseases through speech analysis and that is provided with an acoustic feature quantity extraction means that is not affected by the location where speech is acquired; and a method for activating said inference device.