Subjective Consistency, Series
2022
1'30" latent walk interpolation video
60fps
4K AI upscaled output
Model training: 15000 steps
Machine Learning: StyleGAN2
Model: #subjectiveconsistency
Dataset: the complete documentation of my physical art practice from 2008 until January 2021.
The components of the dataset can be found on cibellecavallibastos.xyz under Bodies of Work/ Physical. It includes sculptures, paintings, text work, ceramics, textiles, watercolours and collages.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Subjective Consistency is a body of work made in cooperation with Machine Learning.
The #subjectiveconsistency model was trained using the complete documentation of my physical art practice from 2008 until January 2021, when the model training had begun.
I select pieces from the generated images and curate the interpolation order. The latent walk is the ML model's decision entirely.
This project intends to investigate a "subjective consistent" practice using Machine Learning and, through human-machine collaborations, attempt to image the aesthetic consistency derived beyond form.
