Key points are not available for this paper at this time.
This paper introduces a computer-based artificial intelligence short video generation system. A stepwise optimization conversion algorithm is presented. This algorithm can realize the fast calculation of large Angle offset and greatly improve the efficiency of short video generation. The complexity of the existing prediction unit is determined by its correlation with time, space and parent neighbor PUs, and the corresponding prediction model is quickly selected according to its complexity, so as to improve the compression efficiency. The system has the functions of content collection to release process, multi-format material synthesis, transcoding processing and so on. The software mainly realizes face recognition, speech recognition, speech synthesis, text recognition, sensitive information recognition, natural language processing and so on. It can better achieve content aggregation, classification and retrieval, thus improving the production effect. The results show that this method has strong functions of learning and intelligent processing. The system can quickly retrieve and organize the text, so as to improve the efficiency of writing, editing and publishing.
Lin et al. (Tue,) studied this question.