Holo-World: Unified Camera, Object and Weather Control for Video World Model

arXiv 2026 65.2 method, benchmark

TLDR

Holo-World unifies camera, object, and weather control in a video world model from a single image, using a dataset and novel adapters.

Reasoning

The paper introduces a novel unified framework for controlling camera, object, and weather in video generation from a single image, with a dedicated dataset and decomposition technique. Strengths include addressing isolated controls and weather transfer; weaknesses include limited scope (first-frame-anchored) and lack of explicit real-world deployment details.

Read-first score

Read-first score 65.2, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 50.

Recency 6%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Citation impact 18%

91.3

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. citation_normalized_percentile=0.91334799

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=baseline,dataset,experiment

Topical relevance 29%

71.4

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=dataset,github

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

FoundationFrontierBridgeMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 386.

Keyword Scores

video world model

world model

generative world model

interactive world model

world simulator

world dynamics prediction

model-based reinforcement learning world model

Deep Analysis

Innovations

First-frame-anchored source-to-state setting for video world models, enabling generation from a single image with explicit camera, object, and optional weather control.
HoloStateData dataset that converts diverse videos into unified control samples for camera, object, and weather supervision.
Unified Scene Adapter that factorizes world preservation and weather transfer into distinct parameter subspaces using rendered background, geometry buffers, and object controls.
Scene-Weather Decomposed CFG (Classifier-Free Guidance) that separately guides scene and weather residuals to strengthen target weather effects without over-amplifying the full condition.

Methodology

The paper proposes Holo-World, a unified controllable video world model. It first builds HoloStateData, a dataset that transforms diverse videos into samples with camera, object, and weather annotations. The model uses a Unified Scene Adapter to maintain controlled scene structure via rendered background, geometry buffers, and object controls while modeling weather-dependent appearance and particle effects. Additionally, Scene-Weather Decomposed CFG guides scene and weather residuals separately to enhance weather effects without distorting the scene.

Key Results

Quantitative and qualitative experiments show that Holo-World maintains precise camera and object control with consistent scene structure while transferring scenes into diverse target weather states, outperforming video-to-video weather editing baselines on weather-state generation.

Limitations

The approach requires explicit camera and object controls as input, which may not always be available in real-world applications.
Weather generation is limited to the weather states present in the HoloStateData dataset, potentially missing rare or extreme conditions.
The method is evaluated only on weather-state generation tasks; its performance on other environmental changes (e.g., lighting, season) is not addressed.

Resources

Paper