Design And Implementation Of Audio And Video Content Retrieval System Based On Artificial Intelligence

The rapid development of Internet technology has stimulated the creative passion of audio and video content producers,and has greatly enriched audio and video content.At the same time,this also puts forward new requirements for the supervision of audio and video content.How to efficiently obtain target information from massive audio and video content has become an important issue for audio and video supervision.In recent years,under the background of the rapid improvement in the performance of AI computing chips,artificial intelligence technologies including machine learning and deep learning have been rapidly developed.Deep learning has made great progress in the fields of computer vision and audio analysis.Based on the above background,this thesis uses computer vision and audio analysis related technologies in deep learning to design a set of audio and video content retrieval system.Based on artificial intelligence analysis.The system is based on the deep learning technology of artificial intelligence,applies different neural network models to realize feature extraction and feature vectorization of audio and video content,and uses high-dimensional retrieval technology to retrieve corresponding content in massive audio and video content by inputting specific information;at the same time;a front-end webpage system and a back-end server with separated front and back ends are designed.Among them,the main neural network models used are InsightFace,CRNN,etc.The neural network model training is mainly implemented by TensorFlow and Keras;the back-end module mainly uses Python as the development language;the front-end uses the Vue framework to complete web development;MySQL is used to complete all the structured data storage work of the system.This thesis firstly elaborates the technical background and practical significance of the subject of determining audio and video content retrieval based on the development of related fields from the 1980s to the present,and summarizes and analyzes the current industry's issue of audio and video content retrieval.On this basis,the research content and objectives of this topic are proposed.Next,the article starts from the front-end and back-end core language,overall system building tools,deep learning core modules,search engines and other dimensions,and introduces the related technologies used,including face recognition algorithms,optical character recognition algorithms,feature extraction,and high-dimensional retrieval Wait.Then it is divided into chapters to describe the specific details of the overall requirements analysis,outline design,functional module division,system design and development,and system testing.