LiZAD: A Lightweight Zero-Shot Anomaly Detection Framework for Industrial Manufacturing

The official implementation of the paper LiZAD: A Lightweight Zero-Shot Anomaly Detection Framework for Industrial Manufacturing accepted at the IEEE International Conference on Omni-Layer Intelligent Systems (COINS 2026).

Abstract

In modern high-throughput industrial production lines, product configurations and visual characteristics frequently change, making it impractical to collect and annotate data for every new scenario. This dynamic setting makes Zero-Shot Anomaly Detection (ZSAD) particularly suitable, as it enables defect detection without requiring training on target-specific samples. Although recent ZSAD approaches show promising results, they are computationally intensive and thus unsuitable for deployment on resource-constrained devices. We propose LiZAD: a first lightweight framework designed for real-time ZSAD specifically tailored for use on edge devices. The proposed approach pairs the dense and spatially aware visual features of DINOv3, crucial for precise pixel-level localization, with the highly computationally efficient text embeddings of MobileCLIP2, mapping them into a shared latent space via low-memory trainable projection heads. Compared to six state- of-the-art ZSAD models, LiZAD achieves an average memory reduction of 61.5%, a parameter reduction of 74.6%, and a speedup of 3.02× in terms of latency. Despite substantial reductions in computational and memory costs, our approach maintains competitive anomaly detection performance, dropping the average P-AUROC by just 6.4% relative to the best state- of-the-art model across the VisA, BTAD, MPDD, and MVTec- AD datasets. Finally, it is successfully deployed on the NVIDIA Jetson NX and Jetson AGX edge devices, and tested on the real production line of the omitted due to double-blind peer review.

Installation

Create a new Conda environment and install the required dependencies:

conda create -n LiZAD python=3.12

conda activate LiZAD

pip install -r requirements.txt

Install MobileCLIP2

To install MobileCLIP2, follow the steps below (see the official MobileCLIP repository for more details).

1. Install MobileCLIP

Clone the MobileCLIP repository and install it:

git clone https://github.com/apple/ml-mobileclip.git
cd ml-mobileclip
pip install .

2. Download the MobileCLIP2-S0 Checkpoint

Download the pretrained MobileCLIP2-S0 checkpoint from Hugging Face:

hf download apple/MobileCLIP2-S0 --local-dir ./pretrained

3. Prepare the Dataset

We follow the standard dataset preprocessing pipeline. The framework requires a meta.json file in the root directory of each dataset.

To generate the required meta.json files, please follow the instructions provided in the AdaCLIP repository.

4. Configure Dataset Paths

Update the base_dir variable in datasets/__init__.py to point to your dataset root directory.

5. Training

Use the following command to train the model:

bash train.sh

6. Evaluation

Use the following command to evaluate the trained model:

bash test.sh

Authors

Uzair Khan¹, Luigi Capogrosso², Muhammad Aqeel¹, Francesco Setti¹, Michele Magno^3,2, Marco Cristani¹

¹ University of Verona

² Interdisciplinary Transformation University of Austria

³ ETH Zurich