mlsanityData to 7 Oct 2026

GICDM: Mitigating Hubness for Reliable Distance-Based Generative Model Evaluation

arXiv 2602.16449 ↗First posted 18 Feb 2026cs.LG cs.AI stat.MLcited by 0

Not on the map. Why?

Abstract (arXiv)

Generative model evaluation commonly relies on high-dimensional embedding spaces to compute distances between samples. We show that dataset representations in these spaces are affected by the hubness phenomenon, which distorts nearest-neighbor relationships and biases distance-based metrics. Building on the classical Iterative Contextual Dissimilarity Measure (ICDM), we introduce Generative ICDM (GICDM), a method to correct neighborhood estimation for both real and generated data. We introduce a multi-scale extension to improve empirical behavior. Extensive experiments on synthetic and real benchmarks demonstrate that GICDM resolves hubness-induced failures, restores reliable metric behavior, and improves alignment with human assessment.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 46 of its 81 references that are in the map. First posted on arXiv 18 Feb 2026.

Builds on

46 of 81 references.

References · 46

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  5. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  6. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  7. 2022Scalable Diffusion Models with Transformerscited by 5,931
  8. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  9. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,736
  10. 2018Large Scale GAN Training for High Fidelity Natural Image Synthesiscited by 2,609
  11. 2017Improved Training of Wasserstein GANscited by 2,223
  12. 2023Consistency Modelscited by 1,778
34 more
  1. 2020Training Generative Adversarial Networks with Limited Datacited by 1,184
  2. 2022MaskGIT: Masked Generative Image Transformercited by 1,082
  3. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 925
  4. 2022Autoregressive Image Generation using Residual Quantizationcited by 706
  5. 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 635
  6. 2021Score-based Generative Modeling in Latent Spacecited by 634
  7. 2016Conditional Image Synthesis With Auxiliary Classifier GANscited by 562
  8. 2023Scaling up GANs for Text-to-Image Synthesiscited by 492
  9. 2022StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasetscited by 489
  10. 2017Word Translation Without Parallel Datacited by 480
  11. 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 425
  12. 2018Assessing Generative Models via Precision and Recallcited by 401
  13. 2022Brain Imaging Generation with Latent Diffusion Modelscited by 267
  14. 2019Residual Flows for Invertible Generative Modelingcited by 256
  15. 2016FMA: A Dataset For Music Analysiscited by 205
  16. 2021Projected GANs Converge Fastercited by 204
  17. 2022Diffusion-GAN: Training GANs with Diffusioncited by 203
  18. 2023Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion modelscited by 199
  19. 2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 193
  20. 2021StyleSwin: Transformer-based GAN for High-resolution Image Generationcited by 152
  21. 2023PFGM++: Unlocking the Potential of Physics-Inspired Generative Modelscited by 73
  22. 2021Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codescited by 71
  23. 2021Data-Efficient Instance Generation from Instance Discriminationcited by 58
  24. 2021Rebooting ACGAN: Auxiliary Classifier GANs with Stable Trainingcited by 57
  25. 2018Metropolis-Hastings Generative Adversarial Networkscited by 40
  26. 2022StudioGAN: A Taxonomy and Benchmark of GANs for Image Synthesiscited by 38
  27. 2022Efficient-VDVAE: Less is morecited by 26
  28. 2022Efficient Image Generation with Variadic Attention Headscited by 17
  29. 2023TopP&R: Robust Support Estimation Approach for Evaluating Fidelity and Diversity in Generative Modelscited by 13
  30. 2025Position: All Current Generative Fidelity and Diversity Metrics are Flawedcited by 12
  31. 2023Probabilistic Precision and Recall Towards Reliable Evaluation of Generative Modelscited by 11
  32. 2023Emergent Asymmetry of Precision and Recall for Measuring Fidelity and Diversity of Generative Models in High Dimensionscited by 5
  33. 2013Fast k-NNG construction with GPU-based quick multi-selectcited by 2
  34. 2025Enhanced Generative Model Evaluation with Clipped Density and Coveragecited by 1

Built on by

Not cited yet.