Another application of our paper "AnyPattern: Towards In-context Image Copy Detection"
The text-to-image model can be used to mimic the style of artwork with little cost, and this threatens the livelihoods and creative rights of artists. To help them protect their work, we treat an artist’s ‘style’ as a ‘pattern’ and generalize the trained pattern retrieval method to identify generated images with style mimicry.
Please refer to the original repository of AnyPattern.
conda create -n style python=3.9
conda activate style
pip install torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 --index-url https://download.pytorch.org/whl/cu118
pip install timm==0.4.12
import requests
import torch
from PIL import Image
from anypattern_style_extractor import preprocessor, create_model
model_name = 'vit_base_pattern'
weight_name = 'vit_ddpmm_8gpu_512_torch2_ap31_pattern.pth.tar'
model = create_model(model_name, weight_name)
url = "https://huggingface.co/datasets/WenhaoWang/AnyPattern/resolve/main/Irises.jpg"
image = Image.open(requests.get(url, stream=True).raw)
x = preprocessor(image).unsqueeze(0)
style_features = model.forward_features(x) # => torch.Size([1, 768])
style_features_normalized = torch.nn.functional.normalize(style_features, p=2, dim=1) # => torch.Size([1, 768])@article{wang2025AnyPattern,
title={AnyPattern: Towards In-context Image Copy Detection},
author={Wang, Wenhao and Sun, Yifan and Tan, Zhentao and Yang, Yi},
booktitle={International Journal of Computer Vision},
year={2025},
}
If you have any questions, feel free to contact Wenhao Wang (wangwenhao0716@gmail.com).
