mlsanityData to 14 Oct 2026

On the Content Bias in Fr'echet Video Distance

arXiv 2404.12391 ↗First posted 18 Apr 2024cs.CV cs.GR cs.LGcited by 52

Not on the map. Why?

Cited by 24 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Fr'echet Video Distance (FVD), a prominent metric for evaluating video generation models, is known to conflict with human perception occasionally. In this paper, we aim to explore the extent of FVD's bias toward per-frame quality over temporal realism and identify its sources. We first quantify the FVD's sensitivity to the temporal axis by decoupling the frame and motion quality and find that the FVD increases only slightly with large temporal corruption. We then analyze the generated videos and show that via careful sampling from a large set of generated videos that do not contain motions, one can drastically decrease FVD without improving the temporal quality. Both studies suggest FVD's bias towards the quality of individual frames. We further observe that the bias can be attributed to the features extracted from a supervised video classifier trained on the content-biased dataset. We show that FVD with features extracted from the recent large-scale self-supervised video models is less biased toward image quality. Finally, we revisit a few real-world examples to validate our hypothesis.

Key facts. Cited by 52 papers in the map (data through 14 Oct 2026). Builds on 80 of its 95 references that are in the map. First posted on arXiv 18 Apr 2024.

Builds on

80 of 95 references.

References · 80

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  6. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  7. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,585
  8. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,355
  9. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  10. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
  11. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
  12. 2017Quo Vadis, Action Recognition? A New Model and the Kinetics Datasetcited by 3,270
38 more
  1. 2016Improved Techniques for Training GANscited by 2,970
  2. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  3. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,581
  4. 2022Video Diffusion Modelscited by 2,305
  5. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,822
  6. 2022Imagen Video: High Definition Video Generation with Diffusion Modelscited by 1,778
  7. 2022Make-A-Video: Text-to-Video Generation without Text-Video Datacited by 1,742
  8. 2021Is Space-Time Attention All You Need for Video Understanding?cited by 1,620
  9. 2023Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Modelscited by 1,476
  10. 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,322
  11. 2018Demystifying MMD GANscited by 1,293
  12. 2018Towards Accurate Generative Models of Video: A New Metric & Challengescited by 1,234
  13. 2021Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrievalcited by 1,177
  14. 2017The "something something" video database for learning and evaluating visual common sensecited by 1,103
  15. 2019FaceForensics++: Learning to Detect Manipulated Facial Imagescited by 1,101
  16. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 944
  17. 2016Temporal Segment Networks: Towards Good Practices for Deep Action Recognitioncited by 902
  18. 2020First Order Motion Model for Image Animationcited by 704
  19. 2023Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video Generatorscited by 670
  20. 2021VideoGPT: Video Generation using VQ-VAE and Transformerscited by 658
  21. 2023Language Model Beats Diffusion -- Tokenizer is Key to Visual Generationcited by 640
  22. 2023VideoMAE V2: Scaling Video Masked Autoencoders with Dual Maskingcited by 528
  23. 2017MoCoGAN: Decomposing Motion and Content for Video Generationcited by 524
  24. 2023VideoPoet: A Large Language Model for Zero-Shot Video Generationcited by 492
  25. 2023DynamiCrafter: Animating Open-domain Images with Video Diffusion Priorscited by 471
  26. 2022MagicVideo: Efficient Video Generation With Latent Diffusion Modelscited by 465
  27. 2015Unsupervised Learning of Video Representations using LSTMscited by 435
  28. 2020Reliable Fidelity and Diversity Metrics for Generative Modelscited by 434
  29. 2018Assessing Generative Models via Precision and Recallcited by 406
  30. 2016Generating Videos with Scene Dynamicscited by 376
  31. 2023LAVIE: High-Quality Video Generation with Cascaded Latent Diffusion Modelscited by 349
  32. 2022MAGVIT: Masked Generative Video Transformercited by 346
  33. 2023VideoFusion: Decomposed Diffusion Models for High-Quality Video Generationcited by 331
  34. 2023Photorealistic Video Generation with Diffusion Modelscited by 306
  35. 2023Show-1: Marrying Pixel and Latent Diffusion Models for Text-to-Video Generationcited by 302
  36. 2021StyleGAN-V: A Continuous Video Generator with the Price, Image Quality and Perks of StyleGAN2cited by 301
  37. 2023Preserve Your Own Correlation: A Noise Prior for Video Diffusion Modelscited by 285
  38. 2021GODIVA: Generating Open-DomaIn Videos from nAtural Descriptionscited by 256

50 most cited of 80 All 80 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 13; 2025: 15; 2026: 24.

52 papers.

Sort:

By area:

  1. 2025Do generative video models understand physical principles?cited by 161
  2. 2025Fast Video Generation with Sliding Tile Attentioncited by 128
  3. 2024EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Video Generationcited by 41
  4. 2024SF-V: Single Forward Video Generation Modelcited by 40
  5. 2024Vivid-ZOO: Multi-View Video Generation with Diffusion Modelcited by 40
  6. 2025DenseDPO: Fine-Grained Temporal Preference Optimization for Video Diffusion Modelscited by 34
  7. 2024Puppet-Master: Scaling Interactive Video Generation as a Motion Prior for Part-Level Dynamicscited by 28
  8. 2024Beyond FVD: Enhanced Evaluation Metrics for Video Generation Qualitycited by 22
  9. 2024Mobile Video Diffusioncited by 18
  10. 2024SurgSora: Object-Aware Diffusion Model for Controllable Surgical Video Generationcited by 15
  11. 2024Fast and Memory-Efficient Video Diffusion Using Streamlined Inferencecited by 14
  12. 2025Frame Guidance: Training-Free Guidance for Frame-Level Control in Video Diffusion Modelscited by 11
38 more
  1. 2025Delving into Latent Spectral Biasing of Video VAEs for Superior Diffusabilitycited by 11
  2. 2025Direct Motion Models for Assessing Generated Videoscited by 10
  3. 2025SwiftVideo: A Unified Framework for Few-Step Video Generation through Trajectory-Distribution Alignmentcited by 10
  4. 2026Generative World Renderercited by 10
  5. 2026DreamID-V:Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformercited by 9
  6. 2026SAW: Toward a Surgical Action World Model via Controllable and Scalable Video Generationcited by 9
  7. 2024Optical Flow Representation Alignment Mamba Diffusion Model for Medical Video Generationcited by 8
  8. 2024Individual Content and Motion Dynamics Preserved Pruning for Video Diffusion Modelscited by 8
  9. 2024Inverse Painting: Reconstructing The Painting Processcited by 6
  10. 2025Stable Cinemetrics : Structured Taxonomy and Evaluation for Professional Video Generationcited by 6
  11. 2026Overcoming Statistical Bias in Action-Controllable World Modelscited by 6
  12. 2025Toward Stable World Models: Measuring and Addressing World Instability in Generative Environmentscited by 5
  13. 2026DreamActor-M2: Universal Character Image Animation via Spatiotemporal In-Context Learningcited by 5
  14. 2025FEAT: Full-Dimensional Efficient Attention Transformer for Medical Video Generationcited by 4
  15. 2025A Survey: Spatiotemporal Consistency in Video Generationcited by 3
  16. 2025FakeParts: a New Family of AI-Generated DeepFakescited by 3
  17. 2026SHIFT: Motion Alignment in Video Diffusion Models with Adversarial Hybrid Fine-Tuningcited by 3
  18. 2026YoCausal: How Far is Video Generation from World Model? A Causality Perspectivecited by 3
  19. 2026World Models: A Comprehensive Survey of Architectures, Methodologies, Reasoning Paradigms, and Applicationscited by 3
  20. 2026V-RAE: Rethinking Video Latent Spaces for Generationcited by 3
  21. 2025GeneVA: A Dataset of Human Annotations for Generative Text to Video Artifactscited by 2
  22. 2025Stable Video-Driven Portraitscited by 2
  23. 2026Measuring 3D Spatial Geometric Consistency in Dynamic Video Generationcited by 2
  24. 2026HairWeaver: Few-Shot Photorealistic Hair Motion Synthesis with Sim-to-Real Guided Video Diffusioncited by 1
  25. 2026SKeDA: A Generative Watermarking Framework for Text-to-video Diffusion Modelscited by 1
  26. 2026ABC: Any-Subset Autoregression via Non-Markovian Diffusion Bridges in Continuous Time and Spacecited by 1
  27. 2026BadDreamer: Transferable Backdoor Attacks against Video World Models for Autonomous Drivingcited by 1
  28. 2026Auditing Generalization in AI-Generated Video Detection: A Six-Control Protocol and the VidAudit Toolkitcited by 1
  29. 2026Physics-Aware Video Generation via Agentic Planning and Graph-Guided Optimizationcited by 1
  30. 2024TC-PDM: Temporally Consistent Patch Diffusion Models for Infrared-to-Visible Video Translationcited by 0
  31. 2024SVS-GAN: Leveraging GANs for Semantic Video Synthesiscited by 0
  32. 2025Physics-Guided Diffusion Priors for Multi-Slice Reconstruction in Scientific Imagingcited by 0
  33. 2026Accelerating Video Inverse Problem Solvers with Autoregressive Diffusion Modelscited by 0
  34. 2026C3VD-DEFCOL: A Deformable Colonoscopy Dataset with Time-Resolved 3D Ground Truth and Realistic Appearancecited by 0
  35. 2026What Do Deepfake Benchmarks Measure? An Audit Using Frozen Self-Supervised Representationscited by 0
  36. 2026UniMoCa: Unifying Motion and Camera Controls as Visual Proxies for Faithful Human Video Generationcited by 0
  37. 2026FlowDance: Music-Driven Dance Video Generation with Parallel Pose and RGB Streamscited by 0
  38. 2026PRISM-Bench: An Audio-Centric Diagnostic Benchmark for Text-to-Audio-Video Generationcited by 0

50 most cited of 52 All 52 (JSON)