Through both the use of print and digital media, Visually Similar Images is a project that explores a feature of Google’s search engine: The ‘Visually Similar Images’ search tool. In June 2011, Google built a feature described as ‘reverse image search’ or ‘content-based image retrieval’. This is when a user uploads an image into Google and the search engine responds by generating image results on account that they are deemed ‘visually similar’.
The result of this technology is what I believe to be an interesting wave in computer-aided curation. The influence that The Internet has had in how the brain reads and responds to imagery is unprecedented in history. Our minds have adapted to online’s immersive, media saturated environment and in many ways we have become desensitised to explicit imagery, cat photos and memes. However, this project won’t contain the aforementioned themes. Or perhaps, by some strange miraculous reason, it will? This publication captures the narratives that play between spontaneous and measured Google searches.
The thing about Google’s ‘Visually Similar Images’ search is that it populates your screen with images based on colour, shape, composition and keywords associated with websites that contain your imputed image. It is these strange coincidences that create a method of curation that is unforeseen. No human could curate art where such happy accidents and anomalies frequently occur. This project explores the beauty of random image generation. Although, the results are not entirely random as they cannot occur without the input of a human. It is the collision of these two minds that brings such interesting results. I see this interaction as a beautiful, harmonious moment where the human and the machine come together at last.