LIFT: Layout-In-Future Video Generation under Large Viewpoint Change via On-Policy Self-Distillation

arXiv 2026 20.1 method

TLDR

Introduces LIFT, an image-to-video generation framework with layout-in-future and camera control, using on-policy self-distillation to learn from sparse last-frame layout annotations.

Reasoning

The paper addresses a practical controllability gap in video generation via explicit future-layout conditioning, supported by a new dataset and experiments. However, the abstract provides limited evidence about limitations and does not frame the work as a world model, making world-model keyword relevance weak.

Read-first score

Read-first score 20.1, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 7.

Recency 5%

100

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

Methodology quality 15%

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

Reproducibility 15%

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

Topical relevance 25%

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

Graph prestige 15%

5.7

PageRank-style prestige over the local citation graph built from available OpenAlex references. local_pagerank=0.0016

Citation impact 15%

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. cited_by_count=0

Citation velocity 10%

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

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 71.

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