现代电影技术Issue(1)：51-57,7.DOI:10.3969/j.issn.1673-3215.2026.01.006
多模态大语言模型辅助电影母版质量检测的方法研究
Research on the method of movie mastering quality inspection assisted by mul-timodal large language model
杨旭1
作者信息
- 1. 中央宣传部电影数字节目管理中心,北京 100866
- 折叠
摘要
Abstract
As the Chinese film industry advances in both quality and efficiency,the traditional manual inspection process,as the core component of movie mastering quality control,has become increasingly prominent.This paper aims at enhanc-ing the quality inspection efficiency for movie mastering,introduces a technical upgrade utilizing multimodal large language model(MLLM)to assist in movie mastering quality inspection.A systematic comparative experiment was conducted to ana-lyze the performance differences between traditional image quality inspection techniques and MLLM-assisted inspection method.The research findings indicate that MLLM-assisted movie mastering quality inspection can substantially improve both efficiency and accuracy.Not only does this research expand the application boundaries of multimodal technology in the field of movie quality inspection,but it also provides new ideas and practical support for promoting the technological in-novation of film industry quality control system.关键词
图像质量评估/多模态大语言模型(MLLM)/电影母版质量检测/人工智能Key words
Image Quality Assessment/Multimodal Large Language Model(MLLM)/Movie Mastering Quality Inspection/Artificial Intelligence分类
信息技术与安全科学引用本文复制引用
杨旭..多模态大语言模型辅助电影母版质量检测的方法研究[J].现代电影技术,2026,(1):51-57,7.
