mlsanityData to 14 Oct 2026

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

arXiv 2312.10634 ↗First posted 17 Dec 2023cs.CV cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 50 of its 62 references that are in the map. First posted on arXiv 17 Dec 2023.

Builds on

50 of 62 references.

References · 50

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  4. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  5. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  6. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  7. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  9. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  10. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  11. 2022Scalable Diffusion Models with Transformerscited by 6,054
  12. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
38 more
  1. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
  2. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,761
  3. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,588
  4. 2022A ConvNet for the 2020scited by 3,579
  5. 2016Improved Techniques for Training GANscited by 2,970
  6. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,626
  7. 2023Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Modelscited by 1,476
  8. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,451
  9. 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,328
  10. 2018Demystifying MMD GANscited by 1,293
  11. 2022MaskGIT: Masked Generative Image Transformercited by 1,099
  12. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  13. 2022Autoregressive Image Generation using Residual Quantizationcited by 722
  14. 2021Score-based Generative Modeling in Latent Spacecited by 636
  15. 2016Conditional Image Synthesis With Auxiliary Classifier GANscited by 565
  16. 2022StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasetscited by 492
  17. 2023Scaling up GANs for Text-to-Image Synthesiscited by 492
  18. 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 434
  19. 2018Assessing Generative Models via Precision and Recallcited by 406
  20. 2017Are GANs Created Equal? A Large-Scale Studycited by 267
  21. 2019Residual Flows for Invertible Generative Modelingcited by 257
  22. 2018Pros and Cons of GAN Evaluation Measurescited by 224
  23. 2022The Role of ImageNet Classes in Fr'echet Inception Distancecited by 213
  24. 2021Projected GANs Converge Fastercited by 208
  25. 2023Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion modelscited by 205
  26. 2019Effectively Unbiased FID and Inception Score and where to find themcited by 165
  27. 2021StyleSwin: Transformer-based GAN for High-resolution Image Generationcited by 152
  28. 2018An empirical study on evaluation metrics of generative adversarial networkscited by 81
  29. 2023PFGM++: Unlocking the Potential of Physics-Inspired Generative Modelscited by 73
  30. 2021Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codescited by 71
  31. 2021Data-Efficient Instance Generation from Instance Discriminationcited by 58
  32. 2021Rebooting ACGAN: Auxiliary Classifier GANs with Stable Trainingcited by 57
  33. 2019LOGAN: Latent Optimisation for Generative Adversarial Networkscited by 42
  34. 2022Rarity Score : A New Metric to Evaluate the Uncommonness of Synthesized Imagescited by 35
  35. 2022Efficient-VDVAE: Less is morecited by 26
  36. 2022Efficient Image Generation with Variadic Attention Headscited by 17
  37. 2022On the Robustness of Quality Measures for GANscited by 11
  38. 2022Curved Representation Space of Vision Transformerscited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Generative Uncertainty in Diffusion Modelscited by 23
  2. 2026Beyond Normal References: Discriminative Few-Shot Anomaly Detectioncited by 0