Anomaly Score: Evaluating Generative Models and Individual Generated Images based on Complexity and Vulnerability

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hwang, Jaehui, Lee, Junghyuk, Lee, Jong-Seok
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916198172590080
author Hwang, Jaehui
Lee, Junghyuk
Lee, Jong-Seok
author_facet Hwang, Jaehui
Lee, Junghyuk
Lee, Jong-Seok
contents With the advancement of generative models, the assessment of generated images becomes more and more important. Previous methods measure distances between features of reference and generated images from trained vision models. In this paper, we conduct an extensive investigation into the relationship between the representation space and input space around generated images. We first propose two measures related to the presence of unnatural elements within images: complexity, which indicates how non-linear the representation space is, and vulnerability, which is related to how easily the extracted feature changes by adversarial input changes. Based on these, we introduce a new metric to evaluating image-generative models called anomaly score (AS). Moreover, we propose AS-i (anomaly score for individual images) that can effectively evaluate generated images individually. Experimental results demonstrate the validity of the proposed approach.
format Preprint
id arxiv_https___arxiv_org_abs_2312_10634
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Anomaly Score: Evaluating Generative Models and Individual Generated Images based on Complexity and Vulnerability
Hwang, Jaehui
Lee, Junghyuk
Lee, Jong-Seok
Computer Vision and Pattern Recognition
Machine Learning
With the advancement of generative models, the assessment of generated images becomes more and more important. Previous methods measure distances between features of reference and generated images from trained vision models. In this paper, we conduct an extensive investigation into the relationship between the representation space and input space around generated images. We first propose two measures related to the presence of unnatural elements within images: complexity, which indicates how non-linear the representation space is, and vulnerability, which is related to how easily the extracted feature changes by adversarial input changes. Based on these, we introduce a new metric to evaluating image-generative models called anomaly score (AS). Moreover, we propose AS-i (anomaly score for individual images) that can effectively evaluate generated images individually. Experimental results demonstrate the validity of the proposed approach.
title Anomaly Score: Evaluating Generative Models and Individual Generated Images based on Complexity and Vulnerability
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2312.10634