license: apache-2.0 library_name: timm tags:

• 图像特征提取
• timm
• transformers

vit_small_patch14_dinov2.lvd142m 模型卡片

基于视觉Transformer（ViT）的图像特征模型。采用自监督DINOv2方法在LVD-142M数据集上预训练。

模型详情

• 模型类型: 图像分类/特征骨干网络
• 模型统计:
  • 参数量（百万）: 22.1
  • 计算量（GMACs）: 46.8
  • 激活值（百万）: 198.8
  • 图像尺寸: 518 × 518
• 相关论文:
  • DINOv2: 无需监督学习鲁棒视觉特征: https://arxiv.org/abs/2304.07193
  • 一张图像等于16x16个词: 大规模图像识别的Transformer: https://arxiv.org/abs/2010.11929v2
• 原始代码库: https://github.com/facebookresearch/dinov2
• 预训练数据集: LVD-142M

模型使用

图像分类

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model('vit_small_patch14_dinov2.lvd142m', pretrained=True)
model = model.eval()

# 获取模型特定的变换（归一化、调整大小）
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(img).unsqueeze(0))  # 将单张图像扩展为批次大小为1

top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5)

图像嵌入

from urllib.request import urlopen
from PIL import Image
import timm

img = Image.open(urlopen(
    'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png'
))

model = timm.create_model(
    'vit_small_patch14_dinov2.lvd142m',
    pretrained=True,
    num_classes=0,  # 移除分类器nn.Linear
)
model = model.eval()

# 获取模型特定的变换（归一化、调整大小）
data_config = timm.data.resolve_model_data_config(model)
transforms = timm.data.create_transform(**data_config, is_training=False)

output = model(transforms(img).unsqueeze(0))  # 输出为(batch_size, num_features)形状的张量

# 或等效方式（无需设置num_classes=0）

output = model.forward_features(transforms(img).unsqueeze(0))
# 输出未池化，为(1, 1370, 384)形状的张量

output = model.forward_head(output, pre_logits=True)
# 输出为(1, num_features)形状的张量

模型比较

在timm的模型结果中探索该模型的数据集和运行时指标。

引用

@misc{oquab2023dinov2,
  title={DINOv2: 无需监督学习鲁棒视觉特征},
  author={Oquab, Maxime and Darcet, Timothée and Moutakanni, Theo and Vo, Huy V. and Szafraniec, Marc and Khalidov, Vasil and Fernandez, Pierre and Haziza, Daniel and Massa, Francisco and El-Nouby, Alaaeldin and Howes, Russell and Huang, Po-Yao and Xu, Hu and Sharma, Vasu and Li, Shang-Wen and Galuba, Wojciech and Rabbat, Mike and Assran, Mido and Ballas, Nicolas and Synnaeve, Gabriel and Misra, Ishan and Jegou, Herve and Mairal, Julien and Labatut, Patrick and Joulin, Armand and Bojanowski, Piotr},
  journal={arXiv:2304.07193},
  year={2023}
}

@article{dosovitskiy2020vit,
  title={一张图像等于16x16个词: 大规模图像识别的Transformer},
  author={Dosovitskiy, Alexey and Beyer, Lucas and Kolesnikov, Alexander and Weissenborn, Dirk and Zhai, Xiaohua and Unterthiner, Thomas and  Dehghani, Mostafa and Minderer, Matthias and Heigold, Georg and Gelly, Sylvain and Uszkoreit, Jakob and Houlsby, Neil},
  journal={ICLR},
  year={2021}
}

@misc{rw2019timm,
  author = {Ross Wightman},
  title = {PyTorch图像模型},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub仓库},
  doi = {10.5281/zenodo.4414861},
  howpublished = {\url{https://github.com/huggingface/pytorch-image-models}}
}