Combined Application of Video Semantic Understanding Technology for Music Video Information Learning

Abstract

Using music practice in computer-aided design can make up for the technical advantages that traditional teaching cannot achieve. It can help teachers and students conduct theoretical research and music creation, integrate teaching resources through modern technological means, and reallocate and effectively utilize music teaching resources. This article mainly studies how to extract effective features from video data with the help of computers to catalog video data. According to the current research situation in this field, the audio features and subtitle features of video are analyzed theoretically. This paper proposes a short video classification and retrieval technology based on computer-assisted image semantic description, and expounds its application in music video information to promote the progress of computer-assisted short video semantic description technology and the development of music video information. The results show that the computer-aided algorithm is more accurate in feature recognition of short video images, 19.99% higher than the comparison algorithm, and can accurately locate the edge contours of short video images.