The Renaissance moved art from a religion-centred to a human-centred way of seeing. This work revisits that shift: instead of replicating classical painting it reduces form, colour and composition to their core relationships — line, light, perspective — and treats that reduction as the material.

A Stable Diffusion model trained on the artist's own visual production, built on ComfyUI, running offline and in real time inside the exhibition space. Nothing is pre-generated; each image belongs to the moment it was made and cannot be reproduced.

A camera set into the screen reads the visitor's body as they step up. From that moment they are no longer an observer: their presence starts the visual process.

Everyday objects in the room — a hat, a bag, a book — become triggers. Every pose, gesture and object continuously reshapes the figure on screen; the body is never fixed.

The result is a temporary, personal digital figure: not a recording, but a momentary collaboration between the visitor's presence and the algorithm.

The same algorithmic logic was also developed through the language of miniature painting, which never relied on linear perspective or realistic proportion: it builds meaning through symbolic scale, flattened space and narrative density.

Digital distortion finds an equivalent in historical abstraction. Where the installation lets the body be rewritten continuously, the miniatures hold that transformation still, as narrative.