# MotionCFG: Boosting Motion Dynamics via Stochastic Concept Perturbation

arXiv: https://arxiv.org/abs/2603.14073 · HTML: https://mlsanity.com/paper/2603.14073/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 31 references that are in the map.
- First posted on arXiv 14 Mar 2026.

## Builds on (most cited first, 29 of 29)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2207.12598/) (2022)
- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets](https://mlsanity.com/paper/2311.15127/) (2023)
- [CLIPScore: A Reference-free Evaluation Metric for Image Captioning](https://mlsanity.com/paper/2104.08718/) (2021)
- [Video Diffusion Models](https://mlsanity.com/paper/2204.03458/) (2022)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [Make-A-Video: Text-to-Video Generation without Text-Video Data](https://mlsanity.com/paper/2209.14792/) (2022)
- [Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval](https://mlsanity.com/paper/2104.00650/) (2021)
- [Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2305.01569/) (2023)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [VideoCrafter1: Open Diffusion Models for High-Quality Video Generation](https://mlsanity.com/paper/2310.19512/) (2023)
- [X-CLIP: End-to-End Multi-grained Contrastive Learning for Video-Text Retrieval](https://mlsanity.com/paper/2207.07285/) (2022)
- [Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models](https://mlsanity.com/paper/2404.07724/) (2024)
- [T2V-CompBench: A Comprehensive Benchmark for Compositional Text-to-video Generation](https://mlsanity.com/paper/2407.14505/) (2024)
- [CFG++: Manifold-constrained Classifier Free Guidance for Diffusion Models](https://mlsanity.com/paper/2406.08070/) (2024)
- [CADS: Unleashing the Diversity of Diffusion Models through Condition-Annealed Sampling](https://mlsanity.com/paper/2310.17347/) (2023)
- [Evaluation of Text-to-Video Generation Models: A Dynamics Perspective](https://mlsanity.com/paper/2407.01094/) (2024)
- [VidGen-1M: A Large-Scale Dataset for Text-to-video Generation](https://mlsanity.com/paper/2408.02629/) (2024)
- [FlowMo: Variance-Based Flow Guidance for Coherent Motion in Video Generation](https://mlsanity.com/paper/2506.01144/) (2025)
- [Minority-Focused Text-to-Image Generation via Prompt Optimization](https://mlsanity.com/paper/2410.07838/) (2024)
- [Frame Guidance: Training-Free Guidance for Frame-Level Control in Video Diffusion Models](https://mlsanity.com/paper/2506.07177/) (2025)
- [Open-World Object Counting in Videos](https://mlsanity.com/paper/2506.15368/) (2025)
- [Improving image synthesis with diffusion-negative sampling](https://mlsanity.com/paper/2411.05473/) (2024)
- [Improving Motion in Image-to-Video Models via Adaptive Low-Pass Guidance](https://mlsanity.com/paper/2506.08456/) (2025)
- [Enhancing Motion in Text-to-Video Generation with Decomposed Encoding and Conditioning](https://mlsanity.com/paper/2410.24219/) (2024)
- [Characterizing Motion Encoding in Video Diffusion Timesteps](https://mlsanity.com/paper/2512.22175/) (2025)

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

- [Rebalancing Reference Frame Dominance to Improve Motion in Image-to-Video Models](https://mlsanity.com/paper/2605.19398/) (2026)

## Abstract (arXiv)

Despite recent advances in Text-to-Video (T2V) synthesis, generating high-fidelity and dynamic motion remains a significant challenge. Existing methods primarily rely on Classifier-Free Guidance (CFG), often with explicit negative prompts (e.g. "static", "blurry"), to suppress undesired artifacts. However, such explicit negations frequently introduce unintended semantic bias and distort object integrity; a phenomenon we define as Content-Motion Drift. To address this, we propose MotionCFG, a framework that enhances motion dynamics by contrasting a target concept with its noise-perturbed counterparts. Specifically, by injecting Gaussian noise into the concept embeddings, MotionCFG creates localized negative anchors that encapsulate a broad complementary space of sub-optimal motion variations. Unlike explicit negations, this approach facilitates implicit hard negative mining without shifting the global semantic identity, allowing for a focused refinement of temporal details. Combined with a piecewise guidance schedule that confines intervention to the early denoising steps, MotionCFG consistently improves motion dynamics across state-of-the-art T2V frameworks with negligible computational overhead and minimal compromise in visual quality. Additionally, we demonstrate that this noise-induced contrastive mechanism is effective not only for sharpening motion trajectories but also for steering complex, non-linear concepts such as precise object numerosity, which are typically difficult to modulate via standard text-based guidance.