Tiny but Trusted: Efficient Vision-Language Reasoning for Time-Series Anomaly Detection
2026-05-28 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors noticed that current vision-language models (VLMs) struggle to find and explain unusual patterns in time-based data. To help with this, they created VisAnomBench, a special benchmark that includes time-series data with detailed explanations of anomalies, gathered using large VLMs. They then fine-tuned a new model called VisAnomReasoner on this benchmark, making it good at spotting and explaining unusual patterns in the data. Tests showed their model is much better than previous ones at accurately finding and describing anomalies, even on different datasets.
Vision-Language ModelsAnomaly DetectionTime-Series DataBenchmarkFine-tuningPrecisionF1 ScoreMultimodal Models
Authors
Xiaona Zhou, Muntasir Wahed, Tianjiao Yu, Constantin Brif, Ismini Lourentzou
Abstract
Recent advances in Vision-Language Models (VLMs) have achieved impressive performance across many tasks, yet prior studies report unsatisfactory performance when applying large language or multimodal models to finding abnormal patterns in sequential data. Public anomaly detection benchmarks typically provide interval annotations but not natural-language rationales, making it difficult to fine-tune VLMs to produce grounded, interpretable decisions. To address this gap, we construct VisAnomBench, a curated benchmark built from public time-series datasets and augmented with high-quality anomaly explanations selected from multiple large VLMs using fine-grained, task-specific rewards. Through fine-tuning on this benchmark, we develop VisAnomReasoner, a parameter-efficient VLM for time-series anomaly detection. Experimental results on VisAnomBench show that VisAnomReasoner achieves more accurate anomaly localization and consistently outperforms all baselines, with improvements of at least 21.23 and 23.87 percentage points in precision and F1, respectively. Additional experiments on the TSB-AD-U benchmark demonstrate strong cross-benchmark generalization, with VisAnomReasoner improving precision and F1 by 9.57 and 13.39 percentage points, respectively.