It seems no place has been left untouched by artificial intelligence. After essentially replacing writers, coders, and creative artists, now AI is coming for yet another object that fuels our social media interactions - the camera lens. Social media without cameras would be a sad and desolate place. However, now with AI one may no longer need a lens that forms the core of the photography experience. Introducing Paragraphica, the world’s first context-to-image camera that visualises a photo based on location data and AI. The device comes as a physical unit and a virtual camera that users can try. What is Paragraphica and how does it work? According to the official website, the camera works by collecting location data with the help of open APIs (application programming interfaces). It uses information such as time of the day, address, weather, and places in the vicinity. Paragraphica writes a paragraph based on the combination of all these details. Later, using the text-to-image AI, the camera converts the paragraph into a photo. Interestingly, the photo is just not any AI rendition, it is according to the website a ‘ complex and nuanced reflection’ of the location the user is at. Or in simpler terms, the image is how the AI sees that place. According to the maker, the photos capture similar moods and tones, but they are not exactly similar to the location. What is the camera made of? The camera comes with a viewfinder that displays real-time descriptions of the current location. Once a user presses the trigger, it will create something known as a scintigraphic (a process that generates scans of the interiors of the human body) interpretation of the description in the paragraph. The camera which bears an unusual look comes with three physical dials that let users control the data and AI parameters to affect the appearance of the image. This is somewhat similar to how traditional cameras with lenses operate. What are the three dials for? According to the website, the first of the three dials on Paragraphica operates similarly to the focal length in an optical lens, however, here it determines the radius in metres of the area the camera searches for data and places. The second dial is instrumental in creating a noise seed that is necessary for AI image diffusion. It comes with values between 0.1 and 1. Meanwhile, the third dial manages the guidance scale - increasing guidance makes the AI perform an in-depth assessment of the paragraph to produce the appropriate image. When it comes to hardware, Paragraphica comes with a system based on Raspberry Pi 4, a large touchscreen, 3D printed housing, and custom electronics. On the software front, Paragraphica is backed by Stable Diffusion API, Python Code, and Noodl.
