近年來,越來越多由人工智慧(Artificial Intelligence, AI)演算生成恐怖影像的方式,以演算法來學習人類視覺所感到驚駭和令人不安的感官條件,並根據數據資料庫運算出怪異反常的影像,已經挑戰了觀看主體的視覺認知。因此,本文將以噩夢機器(Nightmare Machine)、人工靈魂(AI Spirits)演算生成的影像為例,探討恐怖影像的視覺性意涵。本文將透過影像資料蒐集、演算所生產的恐怖影像,放在視覺文化的歷史脈絡底下檢視其意義,考察這種演算法生成的恐怖感,在視覺技術、影像生產層面的延續與差異。並試圖指出,「噩夢機器」和「人工靈魂」透過非單一工作者、非影視劇組的影像生產方式,讓眾人可以透過投稿與投票的網路參與機制,共同累積、增進恐怖感的程度,藉由人工智慧深度學習(Deep Learning)的演算法與人類感知的共作,將集體參與式的影像生產帶入人工智慧演算的恐怖視覺當中,這在以往恐怖影像技術的歷史中是不曾發生的。
In recent years, there have been more and more horror images produced by artificial intelligence. As such, an algorithm is taken for study in the sense of terror and danger acquired through the human eye as well as the calculation of uncommon images based on its data. This has brought challenge to the subjective visual perception. Whether it is the Nightmare Machine or AI Spirits, the algorithm used for image production works out the effect of horror, which on the surface may seem irrelevant superficially, but in fact encompasses a visuality of the non-human. Therefore, this paper discusses how visual technology can affect human beings' way of seeing and the notion of perception from the perspective of visual culture studies. In addition, taking the two cases of AI algorithms, i.e. Nightmare Machine and AI Spirits, as examples, it explores the visual field created via the machine vision within digital datas.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。