Nepotistically Trained Generative-AI Models Collapse

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Bohacek, Matyas, Farid, Hany
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916665276497920
author Bohacek, Matyas
Farid, Hany
author_facet Bohacek, Matyas
Farid, Hany
contents Trained on massive amounts of human-generated content, AI-generated image synthesis is capable of reproducing semantically coherent images that match the visual appearance of its training data. We show that when retrained on even small amounts of their own creation, these generative-AI models produce highly distorted images. We also show that this distortion extends beyond the text prompts used in retraining, and that once affected, the models struggle to fully heal even after retraining on only real images.
format Preprint
id arxiv_https___arxiv_org_abs_2311_12202
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Nepotistically Trained Generative-AI Models Collapse
Bohacek, Matyas
Farid, Hany
Artificial Intelligence
Computer Vision and Pattern Recognition
Trained on massive amounts of human-generated content, AI-generated image synthesis is capable of reproducing semantically coherent images that match the visual appearance of its training data. We show that when retrained on even small amounts of their own creation, these generative-AI models produce highly distorted images. We also show that this distortion extends beyond the text prompts used in retraining, and that once affected, the models struggle to fully heal even after retraining on only real images.
title Nepotistically Trained Generative-AI Models Collapse
topic Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.12202