lensless AI camera converts paragraphs into realistic 'photos'

Paragraphica text-to-image camera by Bjørn Karman

Danish designer Bjørn Karmann has unveiled an innovative camera called Paragraphica, which uses location data and artificial intelligence to generate ‘photos’ of specific places and moments. The lensless camera comes in both physical and virtual forms, allowing users to experience its capabilities firsthand. When looking through the viewfinder, users are presented with a real-time description of their current location. By pressing the trigger, the camera then transforms this description into a scintigraphic representation, effectively creating a unique ‘photo’ of the scene.

using a text-to-image AI, the camera converts the paragraph into a ‘photo’ | all images courtesy of Bjørn Karmann

artificial intelligence meets traditional camera controls

To enhance user control, Paragraphica incorporates physical dials resembling traditional camera controls. These buttons manipulate data and AI parameters, influencing the appearance of the generated photo. By utilizing open APIs, the camera collects data from the surroundings, including address, weather, time, and nearby points of interest. These data points are combined to create a vivid paragraph describing the current location and moment.

The first dial operates in a similar manner to the focal length of an optical lens, but instead of controlling zoom, it adjusts the search radius (in meters) for places and data within the camera’s range. The second dial, reminiscent of film grain, introduces a noise seed into the AI image diffusion process, with values ranging from 0.1 to 1. The third dial controls the guidance scale, influencing the AI’s adherence to the paragraph. Higher values yield a sharper focus, while lower values result in a blurrier interpretation of the photo. In essence, these dials offer users greater control over the camera’s search area, image characteristics, and focus, akin to traditional camera controls.

The resulting ‘photo’ reflects the nuanced essence of the place, as perceived by the AI model. ‘Interestingly the photos do capture some reminiscent moods and emotions from the place but in an uncanny way, as the photos never really look exactly like where I am.’ shares Bjørn Karmann (find more here).

drawing from the star-nosed mole and its finger-like antennae

The design of Paragraphica takes inspiration from the star-nosed mole, a creature that resides and hunts underground, unaffected by light. This unique mammal perceives the world through its finger-like antennae, granting it an extraordinary and intelligent form of “sight.” Serving as a metaphor, the mole showcases the challenge of empathizing with other forms of intelligence and their perception of the world, which can be nearly unimaginable from a human standpoint. Hence, the camera features a prominent bright red antennae in place of the traditional lens.

‘As AI language models are increasingly becoming conscious, we too will have limited imagination of how they will see the world. The camera offers a way of experiencing the world around us, one that is not limited to visual perception alone.’ says Karmann. ‘Through location data and AI image synthesis, Paragraphica provides deeper insight into the essence of a moment through the perspective of other intelligence.’