Method and apparatus for generating speech video

In one embodiment, a method generates a nose average value point, a chin key point, and a lip key point corresponding to a speech input, generates a first person background image by masking a lower end part of a face in an original image using the chin key point and the nose average value point, gen...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: JO SO YEON, SHIN YOON HO, HWANG SUN HEE, LEE SEUNG HYUN, JEON BYOUNG KI, PARK SANG HOON
Format: Patent
Sprache:eng ; kor
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In one embodiment, a method generates a nose average value point, a chin key point, and a lip key point corresponding to a speech input, generates a first person background image by masking a lower end part of a face in an original image using the chin key point and the nose average value point, generates a second person background image by combining the lip key point with the first person background image, and generates a final utterance image from the second person background image. Therefore, the present invention is capable of solving a problem of difficulty in generating the utterance image. 일 실시예는, 음성입력에 해당하는 코 평균값 포인트, 턱 키포인트 및 입술 키포인트를 생성; 상기 턱 키포인트와 상기 코 평균값 포인트를 사용하여, 원본 영상에서 얼굴 하단부를 마스킹하여 제1 인물 배경 영상을 생성; 상기 제1 인물 배경 영상에 상기 입술 키포인트를 합성하여 제2 인물 배경 영상을 생성; 상기 제2 인물 배경 영상으로부터 최종 발화 영상을 생성하는, 방법이다.