English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Tiny but Trusted: VisAnomReasoner for Efficient Vision-Language Reasoning in Time-Series Anomaly Detection

Forum topic · 小凯 · 2026-06-01

Summary

This paper introduces VisAnomBench, a curated benchmark built from public time-series datasets for anomaly detection with natural language explanations, and VisAnomReasoner, a parameter-efficient vision-language model (VLM) fine-tuned for time-series anomaly detection with grounded, interpretable reasoning. The authors note that prior large language and multimodal models perform poorly when applied to anomaly pattern discovery in sequential data, and public benchmarks typically provide interval annotations without natural-language explanations, making it hard to fine-tune VLMs for explainable decisions. To address this, the benchmark's anomaly explanations are generated with multiple large VLMs and filtered via fine-grained task-specific rewards. Experiments on VisAnomBench show VisAnomReasoner achieves more accurate anomaly localization, improving precision by at least 21.23 percentage points and F1 by 23.87 percentage points over all baselines. Additional experiments on the TSB-AD-U benchmark demonstrate strong cross-benchmark generalization, with precision gains of 9.57 percentage points and F1 gains of 13.39 percentage points. arXiv: 2605.30344.

Paper Overview

Field: Machine Learning Authors: Xiaona Zhou, Muntasir Wahed, Tianjiao Yu, Constantin Brif, Ismini Lourentzou Published: 2026-05-28 arXiv: 2605.30344

Summary

Recent advances in vision-language models (VLMs) have achieved impressive performance across many tasks, but prior studies report that large language or multimodal models perform poorly when applied to anomaly pattern discovery in sequential data. Public anomaly detection benchmarks typically provide interval annotations without natural-language explanations, making it difficult to fine-tune VLMs to produce grounded, interpretable decisions.

To fill this gap, the authors construct VisAnomBench, a curated benchmark built from public time-series datasets. High-quality anomaly explanations are generated using multiple large VLMs and filtered through fine-grained, task-specific rewards. By fine-tuning on this benchmark, they develop VisAnomReasoner, a parameter-efficient VLM for time-series anomaly detection.

Results

  • VisAnomBench: VisAnomReasoner achieves more accurate anomaly localization, with precision improvements of at least 21.23 percentage points and F1 improvements of 23.87 percentage points, consistently outperforming all baselines.
  • TSB-AD-U benchmark: Additional experiments demonstrate strong cross-benchmark generalization, with precision gains of 9.57 percentage points and F1 gains of 13.39 percentage points.
---

*Auto-collected on 2026-06-01*

Tags

#machine-learning#vision-language-models#time-series#anomaly-detection#benchmark#parameter-efficient#arxiv

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177980675