# Latent Temporal Discrepancy as Motion Prior: A Loss-Weighting Strategy for Dynamic Fidelity in T2V

arXiv: https://arxiv.org/abs/2601.20504 · HTML: https://mlsanity.com/paper/2601.20504/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 26 references that are in the map.
- First posted on arXiv 28 Jan 2026.

## Builds on (most cited first, 22 of 22)

- [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)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [HunyuanVideo: A Systematic Framework For Large Video Generative Models](https://mlsanity.com/paper/2412.03603/) (2024)
- [ModelScope Text-to-Video Technical Report](https://mlsanity.com/paper/2308.06571/) (2023)
- [VideoCrafter2: Overcoming Data Limitations for High-Quality Video Diffusion Models](https://mlsanity.com/paper/2401.09047/) (2024)
- [VBench-2.0: Advancing Video Generation Benchmark Suite for Intrinsic Faithfulness](https://mlsanity.com/paper/2503.21755/) (2025)
- [Open-Sora Plan: Open-Source Large Video Generation Model](https://mlsanity.com/paper/2412.00131/) (2024)
- [Open-Sora 2.0: Training a Commercial-Level Video Generation Model in 200k](https://mlsanity.com/paper/2503.09642/) (2025)
- [AnimateAnything: Fine-Grained Open Domain Image Animation with Motion Guidance](https://mlsanity.com/paper/2311.12886/) (2023)
- [Follow-Your-Click: Open-domain Regional Image Animation via Short Prompts](https://mlsanity.com/paper/2403.08268/) (2024)
- [VMBench: A Benchmark for Perception-Aligned Video Motion Generation](https://mlsanity.com/paper/2503.10076/) (2025)
- [LivePhoto: Real Image Animation with Text-guided Motion Control](https://mlsanity.com/paper/2312.02928/) (2023)
- [Through-The-Mask: Mask-based Motion Trajectories for Image-to-Video Generation](https://mlsanity.com/paper/2501.03059/) (2025)
- [Optical-Flow Guided Prompt Optimization for Coherent Video Generation](https://mlsanity.com/paper/2411.15540/) (2024)
- [Cinemo: Consistent and Controllable Image Animation with Motion Diffusion Models](https://mlsanity.com/paper/2407.15642/) (2024)
- [MotiF: Making Text Count in Image Animation with Motion Focal Loss](https://mlsanity.com/paper/2412.16153/) (2024)

## Built on by (most cited first, 3 of 3)

- [Omni-WorldBench: Towards a Comprehensive Interaction-Centric Evaluation for World Models](https://mlsanity.com/paper/2603.22212/) (2026)
- [IMPACT: Attention Is the Interaction Map for Scalable Interaction-Aware World Model Training](https://mlsanity.com/paper/2609.00161/) (2026)
- [MotionSpec: Spectral Trajectory Supervision for Motion-Consistent Video Generation](https://mlsanity.com/paper/2609.28095/) (2026)

## Abstract (arXiv)

Video generation models have achieved notable progress in static scenarios, yet their performance in motion video generation remains limited, with quality degrading under drastic dynamic changes. This is due to noise disrupting temporal coherence and increasing the difficulty of learning dynamic regions. {Unfortunately, existing diffusion models rely on static loss for all scenarios, constraining their ability to capture complex dynamics.} To address this issue, we introduce Latent Temporal Discrepancy (LTD) as a motion prior to guide loss weighting. LTD measures frame-to-frame variation in the latent space, assigning larger penalties to regions with higher discrepancy while maintaining regular optimization for stable regions. This motion-aware strategy stabilizes training and enables the model to better reconstruct high-frequency dynamics. Extensive experiments on the general benchmark VBench and the motion-focused VMBench show consistent gains, with our method outperforming strong baselines by 3.31% on VBench and 3.58% on VMBench, achieving significant improvements in motion quality.