Sunday Morning Technology (guangzhou) Co., Ltd.

Zishan Huang

Rongchan Liu

Si Chen

Tian Xia

Qianya Lin

RemyYiYang Ho

The present invention provides a video sticker processing method and device. The method comprises: performing face recognition and speech recognition on video to be processed separately to obtain face position data when face recognition is successful, and speech recognition text when speech recognition is successful; matching speech recognition text with description text of each sticker in sticker library to obtain a target sticker, and obtaining a target video frame according to speech recognition text; adding target sticker to default position or target position of target video frame; wherein, target position is calculated according to the face position data. The present invention can automatically determine the target sticker and its added position according to the face recognition result and the speech recognition result of the video to be processed, realize the intelligent selection and placement of the target sticker, and improve the processing efficiency of the video sticker.

Growing community of inventors

Guangdong, China

Zishan Huang