resnetv2_152x2_bit.goog_teacher_in21k_ft_in1k_384 的模型卡片

一个使用ResNet-V2-BiT（具有预激活的ResNet）的图像分类模型。由论文作者在ImageNet-21k上进行预训练，并在ImageNet-1k上进行微调。

该模型使用以下技术:

• 使用组归一化（GN）与权重标准化（WS）替代批归一化（BN）。

模型详细信息

• 模型类型：图像分类/特征提取器
• 模型统计信息：
  • 参数（百万）：236.3
  • GMACs：136.2
  • 激活值（百万）：132.6
  • 图像尺寸：384 x 384
• 论文：
  • 知识蒸馏：一个好老师温和坚定: https://arxiv.org/abs/2106.05237
  • 大规模迁移（BiT）：通用视觉表示学习: https://arxiv.org/abs/1912.11370
  • 深度残差网络中的恒等映射： https://arxiv.org/abs/1603.05027
• 数据集： ImageNet-1k
• 预训练数据集： ImageNet-21k
• 原始论文： https://github.com/google-research/big_transfer

模型用法

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('resnetv2_152x2_bit.goog_teacher_in21k_ft_in1k_384', pretrained=True)
model = model.eval()

# get model specific transforms (normalization, resize)
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))  # unsqueeze single image into batch of 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(
    'resnetv2_152x2_bit.goog_teacher_in21k_ft_in1k_384',
    pretrained=True,
    features_only=True,
)
model = model.eval()

# get model specific transforms (normalization, resize)
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))  # unsqueeze single image into batch of 1

for o in output:
    # print shape of each feature map in output
    # e.g.:
    #  torch.Size([1, 128, 192, 192])
    #  torch.Size([1, 512, 96, 96])
    #  torch.Size([1, 1024, 48, 48])
    #  torch.Size([1, 2048, 24, 24])
    #  torch.Size([1, 4096, 12, 12])

    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(
    'resnetv2_152x2_bit.goog_teacher_in21k_ft_in1k_384',
    pretrained=True,
    num_classes=0,  # remove classifier nn.Linear
)
model = model.eval()

# get model specific transforms (normalization, resize)
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))  # output is (batch_size, num_features) shaped tensor

# or equivalently (without needing to set num_classes=0)

output = model.forward_features(transforms(img).unsqueeze(0))
# output is unpooled, a (1, 4096, 12, 12) shaped tensor

output = model.forward_head(output, pre_logits=True)
# output is a (1, num_features) shaped tensor

模型比较

在timm中探索该模型的数据集和运行时指标 model results .

引用

@inproceedings{beyer2022knowledge,
  title={Knowledge distillation: A good teacher is patient and consistent},
  author={Beyer, Lucas and Zhai, Xiaohua and Royer, Am{'e}lie and Markeeva, Larisa and Anil, Rohan and Kolesnikov, Alexander},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={10925--10934},
  year={2022}
}

@inproceedings{Kolesnikov2019BigT,
  title={Big Transfer (BiT): General Visual Representation Learning},
  author={Alexander Kolesnikov and Lucas Beyer and Xiaohua Zhai and Joan Puigcerver and Jessica Yung and Sylvain Gelly and Neil Houlsby},
  booktitle={European Conference on Computer Vision},
  year={2019}
}

@article{He2016,
  author = {Kaiming He and Xiangyu Zhang and Shaoqing Ren and Jian Sun},
  title = {Identity Mappings in Deep Residual Networks},
  journal = {arXiv preprint arXiv:1603.05027},
  year = {2016}
}

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