Sora Detector: A Unified Hallucination Detection for Large Text-to-Video Models

TL;DR

The paper introduces SoraDetector, a unified framework for detecting hallucinations in large text-to-video models, improving reliability by evaluating content consistency and using knowledge graphs, with a new benchmark for evaluation.

Contribution

It presents SoraDetector, a novel unified hallucination detection framework for T2V models, incorporating content analysis, knowledge graphs, and automation, along with a new benchmark dataset.

Findings

Effective detection of hallucinations across multiple T2V models

High accuracy in identifying static and dynamic hallucinations

Demonstrated robustness on Sora and other large T2V models

Abstract

The rapid advancement in text-to-video (T2V) generative models has enabled the synthesis of high-fidelity video content guided by textual descriptions. Despite this significant progress, these models are often susceptible to hallucination, generating contents that contradict the input text, which poses a challenge to their reliability and practical deployment. To address this critical issue, we introduce the SoraDetector, a novel unified framework designed to detect hallucinations across diverse large T2V models, including the cutting-edge Sora model. Our framework is built upon a comprehensive analysis of hallucination phenomena, categorizing them based on their manifestation in the video content. Leveraging the state-of-the-art keyframe extraction techniques and multimodal large language models, SoraDetector first evaluates the consistency between extracted video content summary and textual prompts, then constructs static and dynamic knowledge graphs (KGs) from frames to detect hallucination both in single frames and across frames. Sora Detector provides a robust and quantifiable measure of consistency, static and dynamic hallucination. In addition, we have developed the Sora Detector Agent to automate the hallucination detection process and generate a complete video quality report for each input video. Lastly, we present a novel meta-evaluation benchmark, T2VHaluBench, meticulously crafted to facilitate the evaluation of advancements in T2V hallucination detection. Through extensive experiments on videos generated by Sora and other large T2V models, we demonstrate the efficacy of our approach in accurately detecting hallucinations. The code and dataset can be accessed via GitHub.