Digitally Expanded Cinema Through A Computational Vision

本论文在先锋/实验电影制作的历史语境中还原了扩张电影。与以往的扩张电影的叙述不同的是，这种叙述主要是在60年代扩张的艺术环境中的扩张电影，强调移动图像从电影院到画廊的重新定位和媒体融合，而忽视了电影的艺术自主性，这项研究提供了一个以电影为中心的扩张电影概念的重新理论化。本文认为，电影的扩张不仅面向电影的位移和运动图像与其他艺术形式的交叉融合，而且面向计算技术（深度学习，摄影测量，空间计算是本文重点关注的三种技术）渗透到电影纹理中。因此，本研究不仅试图为电影研究（电影史学、先锋理论）做出贡献，而且还试图对当代数字文化和数字美学问题进行批判性反思。

This thesis restores expanded cinema in the historical context of avant-garde/experimentalfilmmaking. Unlike previous accounts of expanded cinema, which mostly situated expanded cinemain '60s expanded arts milieu, stressing moving images' relocation from cinemas to galleries andmedia convergence while ignoring cinema's artistic autonomy, this research provides a cine-centricre-theorisation of the notion of expanded cinema. The cinematic expansion, this thesis argues, notonly faces outwardly in cinema's displacement and the cross-fertilisation between moving imagesand other art forms, but also faces inwardly with computational technologies (deep learning, photogrammetry, spatial computing are the three technologies emphasised in this thesis) permeatingthe filmic texture. Therefore, this research attempts to not only contribute to film studies (filmhistoriography, avant-garde theory) but also provide a critical reflection on contemporary digitalculture and questions concerning digital aesthetics.