@inproceedings{wang-etal-2025-innovative,
title = "Innovative Image Fraud Detection with Cross-Sample Anomaly Analysis: The Power of {LLM}s",
author = "Wang, QiWen and
Yang, Junqi and
Lin, Zhenghao and
Ying, Zhenzhe and
Wang, Weiqiang and
Lin, Chen",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.687/",
doi = "10.18653/v1/2025.acl-long.687",
pages = "14058--14078",
ISBN = "979-8-89176-251-0",
abstract = "The financial industry faces a substantial workload in verifying document images. Existing methods based on visual features struggle to identify fraudulent document images due to the lack of visual clues on the tampering region. This paper proposes CSIAD (Cross-Sample Image Anomaly Detection) by leveraging LLMs to identify logical inconsistencies in similar images. This novel framework accurately detects forged images with slight tampering traces and explains anomaly detection results. Furthermore, we introduce CrossCred, a new benchmark of real-world fraudulent images with fine-grained manual annotations. Experiments demonstrate that CSIAD outperforms state-of-the-art image fraud detection methods by 79.6{\%} (F1) on CrossCred and deployed industrial solutions by 21.7{\%} (F1) on business data. The benchmark is available at https://github.com/XMUDM/CSIAD."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2025-innovative">
<titleInfo>
<title>Innovative Image Fraud Detection with Cross-Sample Anomaly Analysis: The Power of LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">QiWen</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junqi</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhenghao</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhenzhe</namePart>
<namePart type="family">Ying</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weiqiang</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chen</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>The financial industry faces a substantial workload in verifying document images. Existing methods based on visual features struggle to identify fraudulent document images due to the lack of visual clues on the tampering region. This paper proposes CSIAD (Cross-Sample Image Anomaly Detection) by leveraging LLMs to identify logical inconsistencies in similar images. This novel framework accurately detects forged images with slight tampering traces and explains anomaly detection results. Furthermore, we introduce CrossCred, a new benchmark of real-world fraudulent images with fine-grained manual annotations. Experiments demonstrate that CSIAD outperforms state-of-the-art image fraud detection methods by 79.6% (F1) on CrossCred and deployed industrial solutions by 21.7% (F1) on business data. The benchmark is available at https://github.com/XMUDM/CSIAD.</abstract>
<identifier type="citekey">wang-etal-2025-innovative</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.687</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.687/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>14058</start>
<end>14078</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Innovative Image Fraud Detection with Cross-Sample Anomaly Analysis: The Power of LLMs
%A Wang, QiWen
%A Yang, Junqi
%A Lin, Zhenghao
%A Ying, Zhenzhe
%A Wang, Weiqiang
%A Lin, Chen
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F wang-etal-2025-innovative
%X The financial industry faces a substantial workload in verifying document images. Existing methods based on visual features struggle to identify fraudulent document images due to the lack of visual clues on the tampering region. This paper proposes CSIAD (Cross-Sample Image Anomaly Detection) by leveraging LLMs to identify logical inconsistencies in similar images. This novel framework accurately detects forged images with slight tampering traces and explains anomaly detection results. Furthermore, we introduce CrossCred, a new benchmark of real-world fraudulent images with fine-grained manual annotations. Experiments demonstrate that CSIAD outperforms state-of-the-art image fraud detection methods by 79.6% (F1) on CrossCred and deployed industrial solutions by 21.7% (F1) on business data. The benchmark is available at https://github.com/XMUDM/CSIAD.
%R 10.18653/v1/2025.acl-long.687
%U https://aclanthology.org/2025.acl-long.687/
%U https://doi.org/10.18653/v1/2025.acl-long.687
%P 14058-14078
Markdown (Informal)
[Innovative Image Fraud Detection with Cross-Sample Anomaly Analysis: The Power of LLMs](https://aclanthology.org/2025.acl-long.687/) (Wang et al., ACL 2025)
ACL