2022

1'30" video interpolation

HD/60fps

Model training: 11000 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.

The intention of this project is to investigate a "subjective consistent" practice using Machine Learning, and through human-machine collaborations, attempt to image the aesthetic consistency derived beyond form.