- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper proposes VD-DeepStack, a reference-based few-shot anomaly detection (AD) method based on large vision-language models (LVLMs). The key idea is "thinking with visual differences." Previous LVLM-based few-shot AD methods use language chain-of-thought to achieve comparative reasoning. In contrast, VD-DeepStack explicitly conditions language reasoning on query–reference visual differences. Experiments demonstrate the effectiveness of the proposed method.
Critical strengths
- The proposed idea of thinking with visual differences is interesting and technically sound. Compared with existing text-based comparative reasoning, visual-difference-conditioned reasoning makes sense. The motivation is clear.
- This paper is well-organized and easy to follow. Figure 1 clearly illustrates the core idea.
Critical weaknesses
-
The main concern is that the evaluation is not sufficiently comprehensive. (1) Several existing LVLM-based AD methods are not compared in Table 1, such as Anomaly-OV (CVPR 2025 [r1]) and AD-Copilot (reference-based as well) [r2]. (2) Many CLIP-based few-shot AD methods are not compared in Table 1, such as WinCLIP++ and UniVAD [r3]. (3) The backbone LVLMs consider only Qwen-family models. More diverse LVLM backbones should be considered to show generality. (4) The cross-domain evaluation in Table 3 compares only three baselines. This is an important experiment, so more baselines shown in Table 1 should be considered.
-
The value of the reference-based few-shot AD setting is not sufficiently solid. The authors are requested to address this. Specifically, what are its advantages over the zero-shot AD setting (e.g., IAD-R1, Anomaly-OV), which is intuitively more efficient? Furthermore, the performance gap to non-LVLM-based full-shot anomaly detectors is still significant, as shown in Table 1.
-
Figure 3 illustrates only the case where IAD-R1 is wrong, and VD-DeepStack is correct. It is hard to judge their reasoning quality. It should compare the case where both are correct, but their reasoning quality is different.
Given the above three main concerns, my rating at this stage is weak rejection. My final recommendation would depend on the rebuttal.
[r1] Xu, Jiacong, et al. "Towards zero-shot anomaly detection and reasoning with multimodal large language models." CVPR, 2025.
[r2] Jiang, Xi, et al. "AD-Copilot: A Vision-Language Assistant for Industrial Anomaly Detection via Visual In-context Comparison." arXiv preprint arXiv:2603.13779 (2026).
[r3] Gu, Zhaopeng, et al. "Univad: A training-free unified model for few-shot visual anomaly detection." CVPR, 2025.
Questions for authors
In addition to the main weaknesses above, I have the following questions.
-
What exactly is the "Baseline" in Table 2? The number does not match any MLLM reported in Table 1.
-
In Table 2, why does the VisA Prec. number drop significantly after RL?
-
Tables should be placed at the top of the pages.
14 comments
Response to Reviewer emn8, Part 2
Response to Reviewer emn8, Part 2
Question W1(2) — Conventional anomaly detectors
Many CLIP-based few-shot AD methods are not compared in Table 1, such as WinCLIP++ and UniVAD [r3].
Response:
Numbered citations [1]–[11] for this question refer to the references in Part 3.
Due to differences in evaluation paradigms, many existing LVLM-based anomaly detection studies, including IAD-R1 [6], JUDO [7], and Anomaly-R1 [9], do not use conventional anomaly detectors as their primary baselines. Conventional methods typically output anomaly scores and require an empirically selected threshold to obtain normal/abnormal decisions, as discussed in AnomalyGPT [8]. LVLM-based methods instead generate these decisions directly. Nevertheless, we have carefully considered your suggestion and recognize its value, as comparing the two under a common setting can further substantiate our results.
We therefore provide additional comparisons with WinCLIP+ [1], PromptAD [2], InCTRL [3], UniVAD [4], IIPAD [5], and AnomalyDINO [11], all reproduced under a one-shot normal-reference budget. For AnomalyDINO, we evaluate both its original DINOv2 encoder and a frozen DINOv3 replacement, keeping the remaining pipeline unchanged. Regarding FiLo++, we checked the official repository linked in its paper but could not locate a released FiLo++ implementation. The repository currently provides the implementation of its predecessor, FiLo. We therefore do not include FiLo++ in the present comparison.
We evaluate three threshold-selection protocols:
- A: Fixed threshold. A sample is classified as anomalous when its original anomaly score exceeds 0.5, following the fixed-threshold comparison reported in AD-FM [10].
- B: Global threshold. For each method, we select one threshold that maximizes F1 on a labeled 20% subset of the combined test samples and apply it across all four datasets.
- C: Dataset-specific thresholds. For each method and dataset, we select a threshold that maximizes F1 on a labeled 20% subset of that dataset’s test samples.
All methods are evaluated on the full test sets, including the samples used for threshold selection. Protocols B and C therefore provide score-based methods with additional labeled target-domain information. VD-DeepStack uses its directly generated decisions without threshold calibration.
Table 1. Image-level accuracy (%) under a one-shot normal-reference budget. Mean denotes the unweighted average across the four datasets.
| Method | MVTec AD | VisA | MVTec 3D | MPDD | Mean |
|---|---|---|---|---|---|
| (A) Fixed threshold: 0.5 | |||||
| WinCLIP+ | 27.7 | 43.4 | 21.1 | 41.8 | 33.5 |
| PromptAD | 27.7 | 43.4 | 21.1 | 41.8 | 33.5 |
| InCTRL | 58.9 | 57.0 | 42.6 | 50.0 | 52.1 |
| UniVAD | 79.1 | 61.7 | 80.2 | 67.2 | 72.1 |
| IIPAD | 27.7 | 43.4 | 21.1 | 41.8 | 33.5 |
| AnomalyDINO–DINOv2 | 71.4 | 54.0 | 34.5 | 65.8 | 56.4 |
| AnomalyDINO–DINOv3 | 36.9 | 44.8 | 21.2 | 42.9 | 36.4 |
| (B) Global threshold selected using 20% of test labels | |||||
| WinCLIP+ | 77.9 | 64.4 | 77.9 | 57.5 | 69.4 |
| PromptAD | 72.3 | 56.6 | 79.1 | 58.2 | 66.5 |
| InCTRL | 80.5 | 73.9 | 78.5 | 72.1 | 76.2 |
| UniVAD | 88.1 | 81.9 | 81.6 | 77.4 | 82.2 |
| IIPAD | 78.2 | 66.0 | 74.3 | 69.3 | 72.0 |
| AnomalyDINO–DINOv2 | 79.6 | 80.3 | 78.7 | 72.6 | 77.8 |
| AnomalyDINO–DINOv3 | 81.9 | 78.9 | 82.4 | 66.5 | 77.4 |
| (C) Dataset-specific thresholds selected using 20% of test labels | |||||
| WinCLIP+ | 77.8 | 69.5 | 78.9 | 57.5 | 70.9 |
| PromptAD | 72.2 | 57.1 | 79.1 | 57.8 | 66.6 |
| InCTRL | 85.9 | 77.7 | 80.5 | 72.5 | 79.2 |
| UniVAD | 90.4 | 83.3 | 83.1 | 79.6 | 84.1 |
| IIPAD | 78.3 | 69.1 | 78.9 | 71.3 | 74.4 |
| AnomalyDINO–DINOv2 | 83.4 | 80.3 | 81.6 | 67.7 | 78.2 |
| AnomalyDINO–DINOv3 | 89.6 | 78.9 | 83.4 | 62.9 | 78.7 |
| (D) Direct decisions without threshold calibration | |||||
| VD-DeepStack-7B | 91.0 | 82.1 | 76.3 | 78.6 | 82.0 |
| VD-DeepStack-8B | 91.0 | 82.7 | 76.6 | 75.2 | 81.4 |
Reference cited by the reviewer:
[r3] Gu et al. UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detection. CVPR, 2025.
Response to Reviewer emn8, Part 3
Response to Reviewer emn8, Part 3
Question W1(2) — Conventional anomaly detectors (continued)
Many CLIP-based few-shot AD methods are not compared in Table 1, such as WinCLIP++ and UniVAD [r3].
Response:
Threshold selection. Under Protocol A, several conventional detectors perform poorly with a fixed threshold of 0.5. Protocols B and C improve their results by using labeled target data to select a global threshold or separate thresholds for each dataset. Even with this additional information and dataset-specific calibration, both VD-DeepStack variants outperform most compared methods in mean accuracy, including AnomalyDINO–DINOv3.
Obtaining these dataset-specific thresholds in practice requires collecting and labeling target-domain examples. Reusing a threshold across datasets can also reduce accuracy, as shown by transferring thresholds to MVTec 3D while keeping anomaly scores unchanged:
| Method | Own-dataset threshold | Transferred threshold | Drop |
|---|---|---|---|
| IIPAD | 78.92% | 70.22% (from VisA) | 8.70 pp |
| InCTRL | 80.49% | 61.99% (from MVTec AD) | 18.50 pp |
Another point worth noting is that many conventional anomaly detection methods also benefit from potentially privileged settings:
-
PromptAD is retrained for each novel class [2]. It learns new category-specific prompts using that class’s normal samples, producing a separate model for each category. Adapting PromptAD to a single model shared across categories leads to a substantial performance drop, as reported by IIPAD [5].
-
IIPAD partially addresses this limitation by enabling one model to handle multiple categories. However, it first requires few-shot normal samples from all target categories during training [5]. Its one-for-all setting therefore still depends on target-category training: new categories must be incorporated into the training data, rather than handled directly by an unchanged model.
-
UniVAD is a non-end-to-end framework that uses category-specific, empirically set thresholds [4]. Its component-segmentation stage relies on GroundingDINO, where
box_thresholddetermines whether a predicted box is retained, whiletext_thresholddetermines which words are associated with that box. These settings vary across categories: the released pipeline uses box/text thresholds of 0.25/0.25 for bottle and 0.20/0.20 for cable.
VD-DeepStack applies one trained model to new categories using one normal reference, without target-category fine-tuning, category-specific component-segmentation settings, or anomaly-score calibration. It directly produces anomaly decisions, localization, and defect descriptions.
References for W1(2), Parts 2–3:
[1] Jongheon Jeong et al. WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentation. CVPR, 2023.
[2] Xiaofan Li et al. PromptAD: Learning Prompts with only Normal Samples for Few-Shot Anomaly Detection. CVPR, 2024.
[3] Jiawen Zhu et al. Toward Generalist Anomaly Detection via In-context Residual Learning with Few-shot Sample Prompts. CVPR, 2024.
[4] Zhaopeng Gu et al. UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detection. CVPR, 2025.
[5] Wenxi Lv et al. One-for-All Few-Shot Anomaly Detection via Instance-Induced Prompt Learning. ICLR, 2025.
[6] Yanhui Li et al. IAD-R1: Reinforcing Consistent Reasoning in Industrial Anomaly Detection. AAAI, 2026.
[7] Hyunju Kang et al. JUDO: A Juxtaposed Domain-Oriented Multimodal Reasoner for Industrial Anomaly QA. ICLR, 2026.
[8] Zhaopeng Gu et al. AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Models. AAAI, 2024.
[9] Xincheng Yao et al. MMR-AD: A large- scale multimodal dataset for benchmarking general anomaly detection with multimodal large lan- guage models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.
[10] Jingyi Liao et al. AD-FM: Multimodal LLMs for Anomaly Detection via Multi-Stage Reasoning and Fine-Grained Reward Optimization. AAAI, 2026.
[11] Simon Damm et al. AnomalyDINO: Boosting Patch-Based Few-Shot Anomaly Detection with DINOv2. WACV, 2025.
Response to Reviewer emn8, Part 1
Response to Reviewer emn8, Part 1
Question W1(1) — Additional LVLM baselines
Response:
Thank you for this valuable suggestion, which helps make our evaluation more rigorous and comprehensive.
We have added AD-Copilot-7B, which also adopts a reference-based setting, as an additional baseline in the main paper. Its results are presented below alongside our models.
Each entry reports detection accuracy / localization accuracy (%), following Table 1.
Both VD-DeepStack variants achieve higher detection and localization accuracy than AD-Copilot on all four datasets.
Regarding Anomaly-OV, it addresses zero-shot anomaly detection, whereas our work focuses on the one-shot reference-based setting. These are different evaluation paradigms: one-shot methods receive a normal reference that explicitly defines the expected appearance and provides direct visual evidence for identifying defects. Zero-shot methods lack this additional information, placing them at a substantial disadvantage in a direct comparison and making the comparison unfair. Many few-shot studies organize their main comparisons within the same shot setting, as exemplified by PromptAD and UniVAD (please see our response to W2 in Part 5). We follow this practice and prioritize comparisons under the same one-shot normal-reference setting. We therefore have not included Anomaly-OV in this comparison.
References cited by the reviewer in W1:
[r1] Xu, Jiacong, et al. "Towards zero-shot anomaly detection and reasoning with multimodal large language models." CVPR, 2025.
[r2] Jiang, Xi, et al. "AD-Copilot: A Vision-Language Assistant for Industrial Anomaly Detection via Visual In-context Comparison." arXiv preprint arXiv:2603.13779 (2026).