aiqbench analysis
This survey provides a rigorous, assumption-driven taxonomy of multi-modal anomaly detection, moving beyond architectural classifications to address core challenges in defining and separating abnormalities across heterogeneous data sources. It critically evaluates how foundation models are reshaping the field through scalable pretraining and cross-modal transfer while establishing new benchmarks for safety-critical applications like industrial inspection and cybersecurity. The work clarifies open problems regarding robustness and interpretability, offering a clear roadmap for researchers developing adaptive detection systems.
Multi-Modal Anomaly Detection: A Survey
arXiv:2608.24937v1 Announce Type: cross Abstract: Multi-Modal Anomaly Detection (MMAD) detects rare abnormal events from heterogeneous data sources and is increasingly used in safety- and reliability-critical applications such as industrial inspection and cybersecurity. Yet the literature is fragmented across domains and modality combinations, and existing surveys usually group methods by architecture rather than by how abnormality is defined and separated in multi-modal settings. We survey MMAD from an assumption-driven perspective. We formalize the problem, identify five intrinsic characteristics underlying its core challenges, and organize prior work into two complementary paradigms. The first, normality-assumption methods, models regularity via representation learning, cross-modal alignment, and knowledge enhancement. The second, anomaly-assumption methods, sharpens decision boundaries through coarse-grained, structural, and semantic anomaly injection. We also investigate how foundation models are reshaping MMAD through scalable pretraining, flexible cross-modal transfer, and emerging reasoning capabilities. Finally, we compile representative benchmarks and evaluation protocols across domains and highlight open problems and future directions for robust, adaptive, and interpretable MMAD systems.
Abstract:Multi-Modal Anomaly Detection (MMAD) detects rare abnormal events from heterogeneous data sources and is increasingly used in safety- and reliability-critical applications such as industrial inspection and cybersecurity. Yet the literature is fragmented across domains and modality combinations, and existing surveys usually group methods by architecture rather than by how abnormality is defined and separated in multi-modal settings. We survey MMAD from an assumption-driven perspective. We formalize the problem, identify five intrinsic characteristics underlying its core challenges, and organize prior work into two complementary paradigms. The first, normality-assumption methods, models regularity via representation learning, cross-modal alignment, and knowledge enhancement. The second, anomaly-assumption methods, sharpens decision boundaries through coarse-grained, structural, and semantic anomaly injection. We also investigate how foundation models are reshaping MMAD through scalable pretraining, flexible cross-modal transfer, and emerging reasoning capabilities. Finally, we compile representative benchmarks and evaluation protocols across domains and highlight open problems and future directions for robust, adaptive, and interpretable MMAD systems.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2608.24937 [cs.LG]
(or arXiv:2608.24937v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2608.24937 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 250 words
Related coverage
Prompt Engineering in Segment Anything Model: Methodologies, Applications, and Emerging Challenges
arXiv · Aug 18, 2026
OrEdge: Efficient Multi-Modal Anomaly Detection in Distributed Software Systems via Orthogonal-Domain Learning
arXiv · Aug 5, 2026
AI Swiftly Detects and Neutralises 6G Network Threats in Milliseconds
Quantum Zeitgeist · Jun 10, 2026
Hyperspectral Trajectory Image for Multi-Month Trajectory Anomaly Detection
arXiv · Sep 9, 2026