# On the Content Bias in Fr'echet Video Distance

arXiv: https://arxiv.org/abs/2404.12391 · HTML: https://mlsanity.com/paper/2404.12391/

## 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 (most cited first, 30 of 80)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding](https://mlsanity.com/paper/2205.11487/) (2022)
- [A Style-Based Generator Architecture for Generative Adversarial Networks](https://mlsanity.com/paper/1812.04948/) (2018)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [Improved Techniques for Training GANs](https://mlsanity.com/paper/1606.03498/) (2016)
- [Benchmarking Neural Network Robustness to Common Corruptions and Perturbations](https://mlsanity.com/paper/1903.12261/) (2019)
- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [Video Diffusion Models](https://mlsanity.com/paper/2204.03458/) (2022)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [Imagen Video: High Definition Video Generation with Diffusion Models](https://mlsanity.com/paper/2210.02303/) (2022)
- [Make-A-Video: Text-to-Video Generation without Text-Video Data](https://mlsanity.com/paper/2209.14792/) (2022)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2304.08818/) (2023)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [Demystifying MMD GANs](https://mlsanity.com/paper/1801.01401/) (2018)
- [Towards Accurate Generative Models of Video: A New Metric & Challenges](https://mlsanity.com/paper/1812.01717/) (2018)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [FaceForensics++: Learning to Detect Manipulated Facial Images](https://mlsanity.com/paper/1901.08971/) (2019)
- [Improved Precision and Recall Metric for Assessing Generative Models](https://mlsanity.com/paper/1904.06991/) (2019)
- [Temporal Segment Networks: Towards Good Practices for Deep Action Recognition](https://mlsanity.com/paper/1608.00859/) (2016)
- [First Order Motion Model for Image Animation](https://mlsanity.com/paper/2003.00196/) (2020)

## Built on by (most cited first, 30 of 52)

- [Do generative video models understand physical principles?](https://mlsanity.com/paper/2501.09038/) (2025)
- [Fast Video Generation with Sliding Tile Attention](https://mlsanity.com/paper/2502.04507/) (2025)
- [EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Video Generation](https://mlsanity.com/paper/2411.08380/) (2024)
- [SF-V: Single Forward Video Generation Model](https://mlsanity.com/paper/2406.04324/) (2024)
- [Vivid-ZOO: Multi-View Video Generation with Diffusion Model](https://mlsanity.com/paper/2406.08659/) (2024)
- [DenseDPO: Fine-Grained Temporal Preference Optimization for Video Diffusion Models](https://mlsanity.com/paper/2506.03517/) (2025)
- [Puppet-Master: Scaling Interactive Video Generation as a Motion Prior for Part-Level Dynamics](https://mlsanity.com/paper/2408.04631/) (2024)
- [Beyond FVD: Enhanced Evaluation Metrics for Video Generation Quality](https://mlsanity.com/paper/2410.05203/) (2024)
- [Mobile Video Diffusion](https://mlsanity.com/paper/2412.07583/) (2024)
- [SurgSora: Object-Aware Diffusion Model for Controllable Surgical Video Generation](https://mlsanity.com/paper/2412.14018/) (2024)
- [Fast and Memory-Efficient Video Diffusion Using Streamlined Inference](https://mlsanity.com/paper/2411.01171/) (2024)
- [Frame Guidance: Training-Free Guidance for Frame-Level Control in Video Diffusion Models](https://mlsanity.com/paper/2506.07177/) (2025)
- [Delving into Latent Spectral Biasing of Video VAEs for Superior Diffusability](https://mlsanity.com/paper/2512.05394/) (2025)
- [Direct Motion Models for Assessing Generated Videos](https://mlsanity.com/paper/2505.00209/) (2025)
- [SwiftVideo: A Unified Framework for Few-Step Video Generation through Trajectory-Distribution Alignment](https://mlsanity.com/paper/2508.06082/) (2025)
- [Generative World Renderer](https://mlsanity.com/paper/2604.02329/) (2026)
- [DreamID-V:Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer](https://mlsanity.com/paper/2601.01425/) (2026)
- [SAW: Toward a Surgical Action World Model via Controllable and Scalable Video Generation](https://mlsanity.com/paper/2603.13024/) (2026)
- [Optical Flow Representation Alignment Mamba Diffusion Model for Medical Video Generation](https://mlsanity.com/paper/2411.01647/) (2024)
- [Individual Content and Motion Dynamics Preserved Pruning for Video Diffusion Models](https://mlsanity.com/paper/2411.18375/) (2024)
- [Inverse Painting: Reconstructing The Painting Process](https://mlsanity.com/paper/2409.20556/) (2024)
- [Stable Cinemetrics : Structured Taxonomy and Evaluation for Professional Video Generation](https://mlsanity.com/paper/2509.26555/) (2025)
- [Overcoming Statistical Bias in Action-Controllable World Models](https://mlsanity.com/paper/2608.04653/) (2026)
- [Toward Stable World Models: Measuring and Addressing World Instability in Generative Environments](https://mlsanity.com/paper/2503.08122/) (2025)
- [DreamActor-M2: Universal Character Image Animation via Spatiotemporal In-Context Learning](https://mlsanity.com/paper/2601.21716/) (2026)
- [FEAT: Full-Dimensional Efficient Attention Transformer for Medical Video Generation](https://mlsanity.com/paper/2506.04956/) (2025)
- [A Survey: Spatiotemporal Consistency in Video Generation](https://mlsanity.com/paper/2502.17863/) (2025)
- [FakeParts: a New Family of AI-Generated DeepFakes](https://mlsanity.com/paper/2508.21052/) (2025)
- [SHIFT: Motion Alignment in Video Diffusion Models with Adversarial Hybrid Fine-Tuning](https://mlsanity.com/paper/2603.17426/) (2026)
- [YoCausal: How Far is Video Generation from World Model? A Causality Perspective](https://mlsanity.com/paper/2605.30346/) (2026)

## 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.