---
license: cc-by-4.0
language:
- en
pretty_name: VisAnomBench
size_categories:
- 1K<n<10K
task_categories:
- image-text-to-text
tags:
- time-series
- anomaly-detection
- vision-language
- explanations
---

# VisAnomBench

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

**Xiaona Zhou, Muntasir Wahed, Tianjiao Yu, Constantin Brif, Ismini Lourentzou**

### 🎉 Accepted to NeurIPS 2026

[📄 Paper](https://arxiv.org/abs/2605.30344) · [🌐 Project Page](https://plan-lab.github.io/projects/VisAnom/) · [💻 Code](https://github.com/PLAN-Lab/VisAnom) 

[🤗 VisAnomReasoner-3B](https://huggingface.co/PLAN-Lab/VisAnomReasoner-3B) · [🤗 VisAnomReasoner-7B](https://huggingface.co/PLAN-Lab/VisAnomReasoner-7B)

---

## Overview

VisAnomBench is an explanation-augmented benchmark for vision-language time-series anomaly reasoning. Each sample pairs a rendered time-series plot with a prompt and ground-truth anomaly intervals. Training samples also include a step-by-step explanation grounded in the plot. The benchmark is used to train and evaluate [VisAnomReasoner-3B](https://huggingface.co/PLAN-Lab/VisAnomReasoner-3B) and [VisAnomReasoner-7B](https://huggingface.co/PLAN-Lab/VisAnomReasoner-7B).

<p align="center">
  <img src="VisAnomReasoner_teaser.png" width="100%">
</p>


Left: Given a time-series plot, VisAnomReasoner, trained on VisAnomBench, locates anomalies with details grounded in the plot. Right: VisAnomReasoner-7B outperforms the strongest baselines on VisAnomBench and TSB-AD-U.

---

## Dataset Structure

| Split | Samples | UCR-EGI | GutenTAG | UCR-TSAD | KPI |
|---|---:|---:|---:|---:|---:|
| train | 2,576 | 1,682 | 563 | 200 | 131 |
| test | 740 | 415 | 247 | 49 | 29 |

### Data Fields

| Field | Description |
|---|---|
| `image` | Path to the plot image, relative to the root of the [VisAnom code repository](https://github.com/PLAN-Lab/VisAnom), for example `VisAnomBench/VisAnomBench_plot_test/test_0.png`. |
| `query` | Prompt given to the model: context about the series, task instructions, and the required output format. |
| `resp` | Target response: image description, rationale, step-by-step reasoning, and a final answer of the form `<anomaly>True</anomaly><index>(start, end)</index>` followed by the reasoning in `<think>...</think>`. Empty in the test split. |
| `anomalies` | Ground-truth anomaly intervals as inclusive `(start, end)` pairs, for example `(52000, 52620)`. KPI intervals use timestamps, for example `(2016-11-09 03:35:00, 2016-11-09 03:45:00)`. |

## Download

To use the data with the VisAnom code, run this from the root of the code repository:

```bash
hf download PLAN-Lab/VisAnomBench \
  --repo-type dataset \
  --include "VisAnomBench_train.json" \
  --include "VisAnomBench_test.json" \
  --include "VisAnomBench_plot_train/*" \
  --include "VisAnomBench_plot_test/*" \
  --include "VisAnomBench_train_ts/*" \
  --include "VisAnomBench_test_ts/*" \
  --local-dir VisAnomBench
```

## How the Explanations Were Created

For each time-series segment, four general-purpose VLMs (Grok-4-Fast, LLaMA-4-Maverick, Gemma-3-27B-IT, and Qwen2.5-VL-32B) generated candidate outputs, each with an anomaly decision, predicted intervals, and a reasoning trace. Candidates were ranked by a composite reward that combines anomaly localization accuracy with explanation quality (visual groundedness, axis awareness, and clarity) scored by Qwen2.5-VL-72B, and the highest-scoring candidate was kept as the target response. See the paper for details.

## Source Datasets

| Source | Reference | License |
|---|---|---|
| [KPI](https://github.com/NetManAIOps/KPI-Anomaly-Detection) | Xu et al., "Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications," WWW 2018 | MIT |
| [GutenTAG](https://github.com/TimeEval/GutenTAG) | Wenig et al., "TimeEval: A Benchmarking Toolkit for Time Series Anomaly Detection Algorithms," PVLDB 2022 | MIT |
| [UCR-EGI](https://doi.org/10.5441/002/edbt.2020.09) | Gao, Lin, and Brif, "Ensemble Grammar Induction for Detecting Anomalies in Time Series," EDBT 2020 | See source |
| [UCR-TSAD](https://www.cs.ucr.edu/~eamonn/time_series_data_2018/) | Wu and Keogh, "Current Time Series Anomaly Detection Benchmarks are Flawed and are Creating the Illusion of Progress," TKDE 2021 | See source |

The source time series remain subject to the terms of their original sources. Please also cite the source datasets when using VisAnomBench.

## License

VisAnomBench is released under the [CC BY 4.0 license](https://creativecommons.org/licenses/by/4.0/).

## Citation

```bibtex
@inproceedings{zhou2026tiny,
  title={Tiny but Trusted: Efficient Vision-Language Reasoning for Time-Series Anomaly Detection},
  author={Zhou, Xiaona and Wahed, Muntasir and Yu, Tianjiao and Brif, Constantin and Lourentzou, Ismini},
  booktitle={Advances in Neural Information Processing Systems (NeurIPS)},
  year={2026}
}
```