Non-Object Photography is a study of perception and presence, centered on the background, liminal space that usually remains beyond attention. Human visual perception — both in everyday life and in art — tends to focus on objects, figures, narratives. The gaze doesn't capture the image, the landscape as a whole, but glides from object to object. This project proposes to pay attention to what is usually ignored: to the connective tissue of the world, to areas outside the center of the gaze. Using neural networks that programmatically remove recognizable objects, photographs are cleansed of figurative content. As a result, only those elements of the image that resist classification remain: textures, gradients, light, shadow — everything that usually serves as background. These images resemble dreams or memory fragments: impersonal, "de-humanized", plotless and event-less. The project is based on a unique, non-public dataset; at its foundation is the author's archive of more than 40,000 photographs taken in 2023–2025 by the project's author. This is a personal visual chronicle, transformed into an anonymous and non-objective experience.