Illusion or Integrity? Geometrical Consistency Metric for AIGC Video Quality Evaluation

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Yifei Xue, Yuanchen Fei, Hao Zhang, Chenzhi Nie, Tie ji, Yizhen Lao
• Venue: arXiv: cs.CV
• Published: 2026-08-10
• Collected: 2026-08-11
• URL: http://arxiv.org/abs/2608.09594v1
• PDF: https://arxiv.org/pdf/2608.09594v1
• DOI: UNKNOWN
• Registry group: academic_apis_and_preprints
• Evidence role: canonical_academic_index
• Language: en
• Country / region: UNKNOWN

Plain-Language Brief

The paper presents a new benchmark, called GeoCon-Bench, to evaluate the quality of AI-generated videos by measuring their compliance with physical principles and geometric consistency across frames. The benchmark quantitatively measures global motion through translation estimation and homography or fundamental matrix models using background correspondences, reporting metrics like inlier ratio and geometric error.

What the Document Says

• The paper focuses on AI-driven video generation and the need for reliable video quality assessment (VQA) metrics.
• Existing VQA methods evaluate video quality through visual harmony, video-text consistency, and domain-specific alignment but lack quantitative metrics for physical law fidelity.
• The authors introduce GeoCon-Bench to assess AIGC videos based on their compliance with physical principles by measuring geometric consistency across frames.
• GeoCon-Bench captures global motion through translation estimation, fits homography or fundamental matrix models using background correspondences, and reports complementary metrics such as inlier ratio and geometric error.
• The paper includes a dataset of 20 scenes across six motion categories.
• Experiments on state-of-the-art AIGC models demonstrate the reliability of GeoCon-Bench as a VQA metric.

Key Concepts Mentioned

• AI-driven video generation
• video quality assessment (VQA)
• AIGC videos
• physical principles
• geometric consistency
• GeoCon-Bench
• translation estimation
• homography or fundamental matrix models
• inlier ratio
• geometric error
• dataset

Claims Made by the Document

• Existing VQA methods lack quantitative metrics for physical law fidelity.
• GeoCon-Bench is a reliable VQA metric for AIGC videos.

Evidence Used

• The authors mention the need for quantitative metrics for physical law fidelity in existing VQA methods.
• Experiments on state-of-the-art AIGC models demonstrate the reliability of GeoCon-Bench as a VQA metric.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

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.