MotiMotion: Motion-Controlled Video Generation with Visual Reasoning

Hugging Face Daily Papers Papers

Summary

MotiMotion introduces a reasoning-then-generation framework for motion-controlled video generation that uses vision-language reasoning to refine trajectories and a confidence-aware control scheme to improve plausibility, outperforming existing approaches on a new benchmark.

Current motion-controlled image-to-video generation models rigidly follow user-provided trajectories that are often sparse, imprecise, and causally incomplete. Such reliance often yields unnatural or implausible outcomes, especially by missing secondary causal consequences. To address this, we introduce MotiMotion, a novel framework that reformulates motion control as a reasoning-then-generation problem. To encourage causally grounded and commonsense-consistent interactions, we leverage a training-free vision-language reasoner to refine image-space coordinates of primary trajectories and to hallucinate plausible secondary motions. To further improve motion naturalness, we propose a confidence-aware control scheme that modulates guidance strength, enabling the model to closely follow high-confidence plans while correcting artifacts under low-confidence inputs with its internal generative priors. To support systematic evaluation, we curate a new image-to-video benchmark, MotiBench, consisting of interaction-centric scenes where new events are triggered by motion. Both VLM-based evaluation and a human study on MotiBench demonstrate that MotiMotion produces videos with more plausible object behaviors and interaction, and is preferred over existing approaches.
Original Article
View Cached Full Text

Cached at: 05/26/26, 10:46 PM

Paper page - MotiMotion: Motion-Controlled Video Generation with Visual Reasoning

Source: https://huggingface.co/papers/2605.22818

Abstract

MotiMotion introduces a reasoning-then-generation framework for motion-controlled video generation that improves plausibility through vision-language reasoning and confidence-aware control mechanisms.

Current motion-controlled image-to-video generation models rigidly follow user-provided trajectories that are often sparse, imprecise, and causally incomplete. Such reliance often yields unnatural or implausible outcomes, especially by missing secondary causal consequences. To address this, we introduce MotiMotion, a novel framework that reformulatesmotion controlas a reasoning-then-generation problem. To encourage causally grounded and commonsense-consistent interactions, we leverage a training-freevision-language reasonerto refineimage-space coordinatesof primary trajectories and to hallucinate plausiblesecondary motions. To further improve motion naturalness, we propose aconfidence-aware controlscheme that modulatesguidance strength, enabling the model to closely follow high-confidence plans while correcting artifacts under low-confidence inputs with its internalgenerative priors. To support systematic evaluation, we curate a newimage-to-video benchmark,MotiBench, consisting of interaction-centric scenes where new events are triggered by motion. Both VLM-based evaluation and a human study onMotiBenchdemonstrate that MotiMotion produces videos with more plausible object behaviors and interaction, and is preferred over existing approaches.

View arXiv pageView PDFProject pageGitHub2Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.22818 in a model README.md to link it from this page.

Datasets citing this paper1

#### shinying/motibench Preview• Updated5 days ago • 125

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.22818 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

CogOmniControl: Reasoning-Driven Controllable Video Generation via Creative Intent Cognition

Hugging Face Daily Papers

CogOmniControl is a reasoning-driven framework for controllable video generation that uses a specialized vision-language model (CogVLM) trained on anime production data to infer creative intent from sparse conditions, then guides a diffusion-based generator via reinforcement learning, achieving state-of-the-art results on new benchmarks.

OpenCoF: Learning to Reason Through Video Generation

Hugging Face Daily Papers

OpenCoF introduces a reasoning video dataset and a fine-tuned video generation model that improves temporal reasoning through diverse supervision and explicit reasoning tokens, showing significant gains on four video reasoning benchmarks.

Unlocking Complex Visual Generation via Closed-Loop Verified Reasoning

Hugging Face Daily Papers

Introduces CLVR (Closed-Loop Visual Reasoning), a framework that reformulates text-to-image generation from a single-step process into a closed-loop, multi-step visual reasoning approach using a VLM controller and diffusion models, achieving improved performance on compositional prompts.