许可协议: cc-by-nc-4.0 库名称: timm 标签:

• 图像特征提取
• timm
• transformers

模型卡片: convnextv2_pico.fcmae

一个ConvNeXt-V2自监督特征表示模型。采用全卷积掩码自编码器框架(FCMAE)预训练。该模型无预训练头部，仅适用于微调或特征提取。

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(M): 8.6
  • GMACs: 1.4
  • 激活值(M): 6.1
  • 图像尺寸: 224 x 224
• 相关论文:
  • ConvNeXt V2: 通过掩码自编码器协同设计与扩展卷积网络: https://arxiv.org/abs/2301.00808
• 原始实现: https://github.com/facebookresearch/ConvNeXt-V2
• 预训练数据集: ImageNet-1k

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('convnextv2_pico.fcmae', pretrained=True)
model = model.eval()

# 获取模型特定的变换(标准化、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)

特征图提取

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(
    'convnextv2_pico.fcmae',
    pretrained=True,
    features_only=True,
)
model = model.eval()

# 获取模型特定的变换(标准化、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

for o in output:
    # 打印输出中每个特征图的形状
    # 例如:
    #  torch.Size([1, 64, 56, 56])
    #  torch.Size([1, 128, 28, 28])
    #  torch.Size([1, 256, 14, 14])
    #  torch.Size([1, 512, 7, 7])

    print(o.shape)

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(
    'convnextv2_pico.fcmae',
    pretrained=True,
    num_classes=0,  # 移除分类器nn.Linear
)
model = model.eval()

# 获取模型特定的变换(标准化、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, 512, 7, 7)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。

所有计时数据来自RTX 3090上使用AMP的PyTorch 1.13的eager模式。

引用

@article{Woo2023ConvNeXtV2,
  title={ConvNeXt V2: 通过掩码自编码器协同设计与扩展卷积网络},
  author={Sanghyun Woo, Shoubhik Debnath, Ronghang Hu, Xinlei Chen, Zhuang Liu, In So Kweon and Saining Xie},
  year={2023},
  journal={arXiv预印本 arXiv:2301.00808},
}

@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}}
}