Search papers, labs, and topics across Lattice.
This paper introduces GeoCon-Bench, a novel benchmark for evaluating the quality of AI-generated videos by measuring their adherence to physical principles through geometric consistency. By quantitatively assessing global motion and fitting geometric models, the method provides a robust metric that addresses the shortcomings of existing video quality assessment approaches. Experiments reveal that GeoCon-Bench effectively distinguishes the fidelity of AIGC videos, offering a new standard for quality evaluation in the field.
A new benchmark reveals that many AI-generated videos fail to adhere to fundamental physical principles, challenging the reliability of existing quality assessment metrics.
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.
