The Poetry Obscura, an artistic creation that forms an experiential exploration, delves into the intricate interplay of technology, perception, and human experience within the ever-evolving realm of photography. By harnessing advanced AI object detection, this endeavor breathes life into visual narratives on unconventional mediums, exemplified by the use of receipt paper. The Poetry Camera is a compact, interactive device that challenges conventional photography by turning everyday moments into artistic encounters.
The term "Poetry Obscura" captures the unique and poetic quality of the images generated by AI object detection. The camera's understated design, combined with its enigmatic results, creates a fascinating blend of ambiguity and amusement. Everyday objects transform into visual compositions, surpassing simple depiction to embody an artistic expression similar to poetry.
The generation process of images uses OpenCV's object detection to create brief word-based compositions on receipt paper, injecting an element of uncertainty that elevates surprise and creativity in the conventional image creation process. The images produced by the project have a subtle and enigmatic essence that invites viewers to interpret them as poetic verses.
The technical implementation revolves around creating meaning with non-traditional tools. The project utilizes Raspberry Pi, Google Coral TPU, a thermal printer, and a Pi camera to integrate datasets and machine learning techniques into the creative process seamlessly. Exploring alternatives like AI poetry and neural style transfer is crucial due to the inherent uncertainties.
In addition to technical proficiency, the project envisions an interactive installation with an image gallery. The dynamic interaction between the portable camera and its audience, along with visual feedback on receipt paper, creates an immersive experience that challenges conventional photography notions. The project investigates potential interpretations resulting from AI object detection, pushing the limits of conventional image-making methods and offering a new outlook on the fundamental capacities of the medium.
2022. 03 - Ryan Oh