# DiViD: Disentangled Video Diffusion for Static-Dynamic Factorization

arXiv: https://arxiv.org/abs/2507.13934 · HTML: https://mlsanity.com/paper/2507.13934/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Video generation, one of 14 directions in Diffusion models.
- Builds on 24 of its 32 references that are in the map.
- First posted on arXiv 18 Jul 2025.

## Placed in

- Diffusion models / [Video generation](https://mlsanity.com/diffusion/video-generation/)

## Builds on (most cited first, 24 of 24)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Diffusion Models: A Comprehensive Survey of Methods and Applications](https://mlsanity.com/paper/2209.00796/) (2022)
- [InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets](https://mlsanity.com/paper/1606.03657/) (2016)
- [MoCoGAN: Decomposing Motion and Content for Video Generation](https://mlsanity.com/paper/1707.04993/) (2017)
- [Diffusion Autoencoders: Toward a Meaningful and Decodable Representation](https://mlsanity.com/paper/2111.15640/) (2021)
- [Weakly-Supervised Disentanglement Without Compromises](https://mlsanity.com/paper/2002.02886/) (2020)
- [Decomposing Motion and Content for Natural Video Sequence Prediction](https://mlsanity.com/paper/1706.08033/) (2017)
- [A Good Image Generator Is What You Need for High-Resolution Video Synthesis](https://mlsanity.com/paper/2104.15069/) (2021)
- [Disentangled Sequential Autoencoder](https://mlsanity.com/paper/1803.02991/) (2018)
- [Disentangled Representation Learning](https://mlsanity.com/paper/2211.11695/) (2022)
- [Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observations](https://mlsanity.com/paper/1705.08841/) (2017)
- [Unsupervised Learning of Disentangled and Interpretable Representations from Sequential Data](https://mlsanity.com/paper/1709.07902/) (2017)
- [S3VAE: Self-Supervised Sequential VAE for Representation Disentanglement and Data Generation](https://mlsanity.com/paper/2005.11437/) (2020)
- [Disentangling Factors of Variation Using Few Labels](https://mlsanity.com/paper/1905.01258/) (2019)
- [Learning Disentangled Representations in the Imaging Domain](https://mlsanity.com/paper/2108.12043/) (2021)
- [Diffusion Video Autoencoders: Toward Temporally Consistent Face Video Editing via Disentangled Video Encoding](https://mlsanity.com/paper/2212.02802/) (2022)
- [Contrastively Disentangled Sequential Variational Autoencoder](https://mlsanity.com/paper/2110.12091/) (2021)
- [Where and What? Examining Interpretable Disentangled Representations](https://mlsanity.com/paper/2104.05622/) (2021)
- [Commutative Lie Group VAE for Disentanglement Learning](https://mlsanity.com/paper/2106.03375/) (2021)
- [Diffusion Model with Cross Attention as an Inductive Bias for Disentanglement](https://mlsanity.com/paper/2402.09712/) (2024)
- [Disentangled Recurrent Wasserstein Autoencoder](https://mlsanity.com/paper/2101.07496/) (2021)
- [Sample and Predict Your Latent: Modality-free Sequential Disentanglement via Contrastive Estimation](https://mlsanity.com/paper/2305.15924/) (2023)
- [Review of Disentanglement Approaches for Medical Applications -- Towards Solving the Gordian Knot of Generative Models in Healthcare](https://mlsanity.com/paper/2203.11132/) (2022)
- [Sequential Disentanglement by Extracting Static Information From A Single Sequence Element](https://mlsanity.com/paper/2406.18131/) (2024)

## Built on by (most cited first, 1 of 1)

- [The 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real): Methods and Results](https://mlsanity.com/paper/2509.10463/) (2025)

## Abstract (arXiv)

Unsupervised disentanglement of static appearance and dynamic motion in video remains a fundamental challenge, often hindered by information leakage and blurry reconstructions in existing VAE- and GAN-based approaches. We introduce DiViD, the first end-to-end video diffusion framework for explicit static-dynamic factorization. DiViD's sequence encoder extracts a global static token from the first frame and per-frame dynamic tokens, explicitly removing static content from the motion code. Its conditional DDPM decoder incorporates three key inductive biases: a shared-noise schedule for temporal consistency, a time-varying KL-based bottleneck that tightens at early timesteps (compressing static information) and relaxes later (enriching dynamics), and cross-attention that routes the global static token to all frames while keeping dynamic tokens frame-specific. An orthogonality regularizer further prevents residual static-dynamic leakage. We evaluate DiViD on real-world benchmarks using swap-based accuracy and cross-leakage metrics. DiViD outperforms state-of-the-art sequential disentanglement methods: it achieves the highest swap-based joint accuracy, preserves static fidelity while improving dynamic transfer, and reduces average cross-leakage.