wmc_e6 2018.10.22 | more works by Hector Rodriguez from his solo “Hidden Variables”: Theorem 9: Latent Figures (2018) | 羅海德作品繼續介紹：《Theorem 9》（第九定理）的「潛」影像（2018）|
《Theorem 9: Latent Figures》是《Theorem》系列的最新近的衍生作。系列作品均以特定的圖像庫重組其他電影影像。新作以資訊分析中「單體體積最大化」方法選取尚盧．高達《阿爾發城》中最有代表性的畫面，作為演算的圖像庫。圖像庫分成四組，各自透過「非負矩陣分解」將電影中每幀重構，從而組合出各式的視覺類比及比擬，或稱「潛在的形象」。這就是WMC_e6「膨脹電影」(Cinema Expaning) 系列所探索的「象裡有象」的典範罷。
Artist’s on-line archive for Theorem 9 with video demos [… …]
Theorem 9: Latent Figures (2018), a single-channel installation, is the latest entry in the Theorem series. All of the works in the series approximately use a fixed visual dictionary of still images to reconstruct other images. The work uses Simplex Volume Maximization, a data analysis technique, to select highly representative still frames from Jean-Luc Godard’s film Alphaville (1965). These images function as a dictionary.
The dictionary is partitioned into four sets, and every frame in the entire movie is approximately reconstructed using each of these four sets. The reconstructions are achieved using Non-negative Matrix Factorization.
The combination of diverse frames in the dictionary produces visual analogies and metaphors or latent figures. There are always “images within an image,” which is what the Chinese title of this exhibition series literally means (translated into “Cinema Expanding”). In this sense, Theorem 9 is the epitome of the entire series.
