Data to 7 Oct 2026

Assessing Generative Models via Precision and Recall

arXiv 1806.00035 ↗First posted 31 May 2018stat.ML cs.LGcited by 401

Not on the map. Why?

Cited by 99 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Recent advances in generative modeling have led to an increased interest in the study of statistical divergences as means of model comparison. Commonly used evaluation methods, such as the Frechet Inception Distance (FID), correlate well with the perceived quality of samples and are sensitive to mode dropping. However, these metrics are unable to distinguish between different failure cases since they only yield one-dimensional scores. We propose a novel definition of precision and recall for distributions which disentangles the divergence into two separate dimensions. The proposed notion is intuitive, retains desirable properties, and naturally leads to an efficient algorithm that can be used to evaluate generative models. We relate this notion to total variation as well as to recent evaluation metrics such as Inception Score and FID. To demonstrate the practical utility of the proposed approach we perform an empirical study on several variants of Generative Adversarial Networks and Variational Autoencoders. In an extensive set of experiments we show that the proposed metric is able to disentangle the quality of generated samples from the coverage of the target distribution.

Key facts. Cited by 401 papers in the map (data through 7 Oct 2026). First posted on arXiv 31 May 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2020: 1; 2021: 54; 2022: 59; 2023: 87; 2024: 72; 2025: 60; 2026: 68.

401 papers.

Sort:

By area:

2020Improved Techniques for Training Score-Based Generative Modelscited by 1,036
2021Tackling the Generative Learning Trilemma with Denoising Diffusion GANscited by 515
2021MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Frontierscited by 466
20233DShape2VecSet: A 3D Shape Representation for Neural Fields and Generative Diffusion Modelscited by 449
2023Finite Scalar Quantization: VQ-VAE Made Simplecited by 425
2022The Vendi Score: A Diversity Evaluation Metric for Machine Learningcited by 299
20223D Neural Field Generation using Triplane Diffusioncited by 274
2021Large Scale Image Completion via Co-Modulated Generative Adversarial Networkscited by 219
2022The Role of ImageNet Classes in Fr'echet Inception Distancecited by 211
2021Projected GANs Converge Fastercited by 204
2023Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion modelscited by 199
2023VIEScore: Towards Explainable Metrics for Conditional Image Synthesis Evaluationcited by 196

38 more

2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 193
2023Selective Amnesia: A Continual Learning Approach to Forgetting in Deep Generative Modelscited by 175
2023Raising the Cost of Malicious AI-Powered Image Editingcited by 160
2022GENIE: Higher-Order Denoising Diffusion Solverscited by 127
2021Pros and Cons of GAN Evaluation Measures: New Developmentscited by 121
2021DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networkscited by 109
2024Why Tabular Foundation Models Should Be a Research Prioritycited by 103
2023On the Stability of Iterative Retraining of Generative Models on their own Datacited by 98
2023Differentially Private Diffusion Models Generate Useful Synthetic Imagescited by 85
2023LLMScore: Unveiling the Power of Large Language Models in Text-to-Image Synthesis Evaluationcited by 85
2023KoLA: Carefully Benchmarking World Knowledge of Large Language Modelscited by 85
2023GIVT: Generative Infinite-Vocabulary Transformerscited by 80
2022Generating High Fidelity Data from Low-density Regions using Diffusion Modelscited by 78
2024Distilling Diffusion Models into Conditional GANscited by 78
2024Seamless Human Motion Composition with Blended Positional Encodingscited by 77
2022Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesiscited by 75
2024A Survey of Multimodal Large Language Model from A Data-centric Perspectivecited by 73
2022Any-resolution Training for High-resolution Image Synthesiscited by 72
2021Unleashing Transformers: Parallel Token Prediction with Discrete Absorbing Diffusion for Fast High-Resolution Image Generation from Vector-Quantized Codescited by 71
2021Aligning Latent and Image Spaces to Connect the Unconnectablecited by 70
2022Synthetic Data -- what, why and how?cited by 66
2021MG-GAN: A Multi-Generator Model Preventing Out-of-Distribution Samples in Pedestrian Trajectory Predictioncited by 61
2022MoDi: Unconditional Motion Synthesis from Diverse Datacited by 59
2021Rebooting ACGAN: Auxiliary Classifier GANs with Stable Trainingcited by 57
2021Adversarial Text-to-Image Synthesis: A Reviewcited by 52
2023Curated LLM: Synergy of LLMs and Data Curation for tabular augmentation in low-data regimescited by 52
2024Erasing Undesirable Influence in Diffusion Modelscited by 52
2021On Aliased Resizing and Surprising Subtleties in GAN Evaluationcited by 51
2024On the Content Bias in Fr'echet Video Distancecited by 51
2023Class-Balancing Diffusion Modelscited by 47
2024JetFormer: An Autoregressive Generative Model of Raw Images and Textcited by 44
2022Can Push-forward Generative Models Fit Multimodal Distributions?cited by 41
2022StudioGAN: A Taxonomy and Benchmark of GANs for Image Synthesiscited by 38
2022Show Me What and Tell Me How: Video Synthesis via Multimodal Conditioningcited by 37
2023Unlearnable Examples for Diffusion Models: Protect Data from Unauthorized Exploitationcited by 37
2022Mutual Information Divergence: A Unified Metric for Multimodal Generative Modelscited by 35
2024Consistency-diversity-realism Pareto fronts of conditional image generative modelscited by 35
2022When do GANs replicate? On the choice of dataset sizecited by 34

50 most cited of 401 All 401 (JSON)