tf_efficientnet_b2.aa_in1k 模型卡片

一种基于 EfficientNet 的图片分类模型。在 ImageNet-1k 数据集上使用 Tensorflow 和 paper 作者自动数据增强训练得到，由 Ross Wightman 移植到 PyTorch。

模型细节

• 模型类型: 图片分类 / 特征提取主干
• 模型统计信息:
  • 参数 (M): 9.1
  • GMACs: 1.0
  • 激活量 (M): 13.8
  • 图片尺寸: 260 x 260
• 论文:
  • EfficientNet: 重新思考卷积神经网络的模型缩放: https://arxiv.org/abs/1905.11946
  • AutoAugment: 从数据中学习数据增强策略: https://arxiv.org/abs/1805.09501
• 数据集: ImageNet-1k
• 原始来源: https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet

模型用法

图片分类

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('tf_efficientnet_b2.aa_in1k', 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(
    'tf_efficientnet_b2.aa_in1k',
    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, 16, 130, 130])
    #  torch.Size([1, 24, 65, 65])
    #  torch.Size([1, 48, 33, 33])
    #  torch.Size([1, 120, 17, 17])
    #  torch.Size([1, 352, 9, 9])

    print(o.shape)

图片嵌入

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(
    'tf_efficientnet_b2.aa_in1k',
    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, 1408, 9, 9) 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度量。 model results

引用

@inproceedings{tan2019efficientnet,
  title={Efficientnet: Rethinking model scaling for convolutional neural networks},
  author={Tan, Mingxing and Le, Quoc},
  booktitle={International conference on machine learning},
  pages={6105--6114},
  year={2019},
  organization={PMLR}
}

@inproceedings{47890,
    title       = {AutoAugment: Learning Augmentation Policies from Data},
    author      = {Ekin Dogus Cubuk and Barret Zoph and Dandelion Mane and Vijay Vasudevan and Quoc V. Le},
    year        = {2019},
    URL = {https://arxiv.org/pdf/1805.09501.pdf}
}

@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}}
}