摘要 |
<p>PROBLEM TO BE SOLVED: To provide a voice input device configured to generate a high-quality voice signal even when a speech of a person wearing a mask is input.SOLUTION: A voice input device 11 for converting an input speech to a voice signal includes: an image processing section 20 which acquires an image of a person captured, and extracts a mouth part as a mouth image; a determination section 30 which detects a mask from the mouth image; and a voice processing section 40 which acquires a collected speech and increases gain of a predetermined frequency band in the speech when the determination section 30 detects the mask.</p> |