PanFlow: Decoupled Motion Control for Panoramic Video Generation

arXiv 2025 27.3 method, benchmark, application

TLDR

PanFlow decouples camera rotation from optical flow to control large motions in panoramic video generation, adding spherical noise warping and a motion-rich dataset.

Reasoning

Strengths include explicit motion control for panoramic video, a novel spherical noise warping strategy, and a large-scale annotated dataset with strong empirical results. Weaknesses are that the work focuses narrowly on panoramic video generation and makes no connection to world models or dynamics prediction.

Read-first score

Read-first score 27.3, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 1.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Methodology quality 25%

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

Reproducibility 25%

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

Topical relevance 42%

1.4

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

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 282.

Keyword Scores

video world model

world model

world simulator

generative world model

interactive world model

world dynamics prediction

model-based reinforcement learning world model

Resources

Paper