dorsal/arxiv
View SchemaCascading multi-agent anomaly detection in surveillance systems via vision-language models and embedding-based classification
| Authors | Tayyab Rehman, Giovanni De Gasperis, Aly Shmahell |
|---|---|
| Categories | |
| ArXiv ID | 2601.06204vv3 |
| URL | https://arxiv.org/abs/2601.06204 |
| License | http://creativecommons.org/licenses/by-nc-nd/4.0/ |
Abstract
Intelligent anomaly detection in dynamic visual environments requires reconciling real-time performance with semantic interpretability. Conventional approaches address only fragments of this challenge. Reconstruction-based models capture low-level deviations without contextual reasoning, object detectors provide speed but limited semantics, and large vision-language systems deliver interpretability at prohibitive computational cost. This work introduces a cascading multi-agent framework that unifies these complementary paradigms into a coherent and interpretable architecture. Early modules perform reconstruction-gated filtering and object-level assessment, while higher-level reasoning agents are selectively invoked to interpret semantically ambiguous events. The system employs adaptive escalation thresholds and a publish-subscribe communication backbone, enabling asynchronous coordination and scalable deployment across heterogeneous hardware. Extensive evaluation on large-scale monitoring data demonstrates that the proposed cascade achieves a threefold reduction in latency compared to direct vision-language inference, while maintaining high perceptual fidelity (PSNR = 38.3 dB, SSIM = 0.965) and consistent semantic labeling. The framework advances beyond conventional detection pipelines by combining early-exit efficiency, adaptive multi-agent reasoning, and explainable anomaly attribution, establishing a reproducible and energy-efficient foundation for scalable intelligent visual monitoring.
{
"annotation_id": "47fa284c-9ed8-45e5-a5ad-afa6d497ab71",
"date_created": "2026-02-17T05:53:08.753000Z",
"date_modified": "2026-02-17T05:53:08.753000Z",
"file_hash": "5f0bc138e056adc54cd7f2e0e98402cfbe4ecab5cb79b41a3538fc417af3aeaa",
"private": false,
"record": {
"abstract": "Intelligent anomaly detection in dynamic visual environments requires reconciling real-time performance with semantic interpretability. Conventional approaches address only fragments of this challenge. Reconstruction-based models capture low-level deviations without contextual reasoning, object detectors provide speed but limited semantics, and large vision-language systems deliver interpretability at prohibitive computational cost. This work introduces a cascading multi-agent framework that unifies these complementary paradigms into a coherent and interpretable architecture. Early modules perform reconstruction-gated filtering and object-level assessment, while higher-level reasoning agents are selectively invoked to interpret semantically ambiguous events. The system employs adaptive escalation thresholds and a publish-subscribe communication backbone, enabling asynchronous coordination and scalable deployment across heterogeneous hardware. Extensive evaluation on large-scale monitoring data demonstrates that the proposed cascade achieves a threefold reduction in latency compared to direct vision-language inference, while maintaining high perceptual fidelity (PSNR = 38.3 dB, SSIM = 0.965) and consistent semantic labeling. The framework advances beyond conventional detection pipelines by combining early-exit efficiency, adaptive multi-agent reasoning, and explainable anomaly attribution, establishing a reproducible and energy-efficient foundation for scalable intelligent visual monitoring.",
"arxiv_id": "2601.06204",
"authors": [
"Tayyab Rehman",
"Giovanni De Gasperis",
"Aly Shmahell"
],
"categories": [
"cs.CV",
"cs.MA"
],
"license": "http://creativecommons.org/licenses/by-nc-nd/4.0/",
"title": "Cascading multi-agent anomaly detection in surveillance systems via vision-language models and embedding-based classification",
"url": "https://arxiv.org/abs/2601.06204",
"version": "v3"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "66ad0f08-534b-43f1-8643-b6405749c8c7",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}