Research on Artificial Intelligence-Enabled Animation Script Creation

Abstract

With the development of artificial intelligence technology, animation script creation has ushered in new opportunities. This study investigates the utilization of an algorithm based on LSTM and attention mechanisms to enhance the automation level of animation script creation. By integrating the advantages of LSTM in capturing temporal sequence dependencies and the superiority of the attention mechanism in processing long sequences, a novel script generation algorithm is proposed. The research results demonstrate significant advantages of this algorithm in terms of coherence, emotional expression, grammatical correctness, and plot coherence of the generated script content. Both in terms of BLEU score and ROUGE score, the algorithm performs exceptionally well after initial iterations and multiple iterations, reaching 0.50 and 0.55 respectively. In comparative experiments regarding grammatical errors, plot jumps, and other aspects, the sentences generated by the research model algorithm significantly outperform other algorithms, exhibiting higher accuracy and coherence. Overall, the algorithm based on LSTM and attention mechanisms can effectively improve the quality of animation script creation, providing new methods and ideas for intelligent script generation, which holds significant theoretical and practical implications.