Problem
Research questions and friction points this paper is trying to address.
video quality assessment
AIGC
geometric consistency
physical plausibility
AI-generated video
Illusion or Integrity? Geometrical Consistency Metric for AIGC Video Quality Evaluation
🤖 AI Summary
Current evaluation methods for AI-generated videos lack quantitative metrics to assess fidelity to physical laws, making it difficult to measure geometric and motion-level realism. To address this gap, this work proposes GeoCon-Bench, a novel benchmark that introduces cross-frame geometric consistency as a core evaluation dimension. The framework quantifies physical plausibility through feature matching, homography or fundamental matrix estimation, and analysis of inlier ratios and geometric errors. Accompanying this benchmark, the authors release a dedicated dataset comprising six motion categories across 20 diverse scenes. Extensive experiments on state-of-the-art AIGC video models demonstrate the effectiveness of GeoCon-Bench, offering an interpretable and quantitative foundation for guiding the optimization of generative models toward greater physical realism.
📝 Abstract
Recently, AI-driven video generation has attracted considerable attention. This surge increases the demand for reliable video quality assessment (VQA) metrics to evaluate AI-generated content (AIGC) videos and guide model optimization. Existing studies assess video quality through visual harmony, video-text consistency, and domain-specific alignment, yet lack quantitative metrics for measuring fidelity to physical laws. To address this limitation, we present a novel benchmark that evaluates the quality of AIGC videos based on their compliance with physical principles by quantitatively measuring geometric consistency across frames extracted from generated sequences. This serves as a proxy for estimating the extent to which generated videos conform to real-world physical rules. Specifically, GeoCon-Bench captures global motion through translation estimation, fits homography or fundamental matrix models using background correspondences, and reports complementary metrics, including inlier ratio and geometric error. We also release a dataset containing 20 scenes across six motion categories. Experiments on state-of-the-art AIGC models demonstrate the reliability of GeoCon-Bench as a video quality assessment metric.
Innovation
Methods, ideas, or system contributions that make the work stand out.
geometric consistency
video quality assessment
AIGC
physical plausibility
homography estimation
🔎 Similar Papers
