数智化“生成式”电影多模态交互构建研究  

Research on Construction of Multi-modal Interaction in “Generative” Films in the Era of Digital Intelligence

作　　者：程樯 陈微 Cheng Qiang;Chen Wei

机构地区：[1]北京电影学院视听传媒学院

出　　处：《电影艺术》2024年第3期110-119,共10页Film Art

摘　　要：数智时代,人工智能正改变着人类与影像的关系,电影作者主体性地位受到挑战。人工智能时代,“生成式”电影制作的发展趋势与AIGC介入电影的边界问题可围绕可见的物理层面、可感的心理层面、可知的动态美学三个层面进行展开,它涵盖多模态电影数智化界域构成、垂直领域模型个性化机制的演绎、人工智能生成与电影虚拟制作的交互策略等问题。“向弱势生成”的人机共创模式为电影创作提供了一种实践范式,进而探索未来人工智能技术在电影领域中新的可能性。In the era of digital intelligence,artificial intelligence(AI)as a core technology is transforming the relationship between human and image,thereby challenging filmmakers as authors.This paper explores the development trends of“generative”film production in the AI era,meanwhile reflects AIGC’s intervention in film-making,focusing on its physical,psychological,and aesthetic aspects.This study also examines the digital intelligence composition of multi-modal AI films,the personalized mechanisms of vertical field models’progression,and the interactive strategies of Artificial Intelligence Generation and virtual film production.The“generation towards disadvantaged”collaborative model between human and machine provides a practical paradigm for film creation,thus exploring new possibilities for AI technology in the film industry.

关 键 词：生成式电影 多模态人工智能 数智化 AIGC 

分 类 号：J91[艺术—电影电视艺术] J93[自动化与计算机技术—控制理论与控制工程] TP18[自动化与计算机技术—控制科学与工程]