Conditional Hallucinations for Image Compression

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Aczel, Till, Wattenhofer, Roger
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910859814502400
author Aczel, Till
Wattenhofer, Roger
author_facet Aczel, Till
Wattenhofer, Roger
contents In lossy image compression, models face the challenge of either hallucinating details or generating out-of-distribution samples due to the information bottleneck. This implies that at times, introducing hallucinations is necessary to generate in-distribution samples. The optimal level of hallucination varies depending on image content, as humans are sensitive to small changes that alter the semantic meaning. We propose a novel compression method that dynamically balances the degree of hallucination based on content. We collect data and train a model to predict user preferences on hallucinations. By using this prediction to adjust the perceptual weight in the reconstruction loss, we develop a Conditionally Hallucinating compression model (ConHa) that outperforms state-of-the-art image compression methods. Code and images are available at https://polybox.ethz.ch/index.php/s/owS1k5JYs4KD4TA.
format Preprint
id arxiv_https___arxiv_org_abs_2410_19493
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Conditional Hallucinations for Image Compression
Aczel, Till
Wattenhofer, Roger
Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
In lossy image compression, models face the challenge of either hallucinating details or generating out-of-distribution samples due to the information bottleneck. This implies that at times, introducing hallucinations is necessary to generate in-distribution samples. The optimal level of hallucination varies depending on image content, as humans are sensitive to small changes that alter the semantic meaning. We propose a novel compression method that dynamically balances the degree of hallucination based on content. We collect data and train a model to predict user preferences on hallucinations. By using this prediction to adjust the perceptual weight in the reconstruction loss, we develop a Conditionally Hallucinating compression model (ConHa) that outperforms state-of-the-art image compression methods. Code and images are available at https://polybox.ethz.ch/index.php/s/owS1k5JYs4KD4TA.
title Conditional Hallucinations for Image Compression
topic Image and Video Processing
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2410.19493