Model card for eca_nfnet_l2.ra3_in1k

A ECA-NFNet-Lite (Lightweight NFNet w/ ECA attention) image classification model. Trained in timm by Ross Wightman.

Normalization Free Networks are (pre-activation) ResNet-like models without any normalization layers. Instead of Batch Normalization or alternatives, they use Scaled Weight Standardization and specifically placed scalar gains in residual path and at non-linearities based on signal propagation analysis.

Lightweight NFNets are timm specific variants that reduce the SE and bottleneck ratio from 0.5 -> 0.25 (reducing widths) and use a smaller group size while maintaining the same depth. SiLU activations used instead of GELU.

This NFNet variant also uses ECA (Efficient Channel Attention) instead of SE (Squeeze-and-Excitation).

Model Details

• Model Type: Image classification / feature backbone
• Model Stats:
  • Params (M): 56.7
  • GMACs: 21.0
  • Activations (M): 47.4
  • Image size: train = 320 x 320, test = 384 x 384
• Papers:
  • High-Performance Large-Scale Image Recognition Without Normalization: https://arxiv.org/abs/2102.06171
  • Characterizing signal propagation to close the performance gap in unnormalized ResNets: https://arxiv.org/abs/2101.08692
• Original: https://github.com/huggingface/pytorch-image-models
• Dataset: ImageNet-1k

Model Usage

Image Classification

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('eca_nfnet_l2.ra3_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)

Feature Map Extraction

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(
    'eca_nfnet_l2.ra3_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, 64, 160, 160])
    #  torch.Size([1, 256, 80, 80])
    #  torch.Size([1, 512, 40, 40])
    #  torch.Size([1, 1536, 20, 20])
    #  torch.Size([1, 3072, 10, 10])

    print(o.shape)

Image Embeddings

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(
    'eca_nfnet_l2.ra3_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, 3072, 10, 10) shaped tensor

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

Model Comparison

Explore the dataset and runtime metrics of this model in timm model results.

Citation

@article{brock2021high,
  author={Andrew Brock and Soham De and Samuel L. Smith and Karen Simonyan},
  title={High-Performance Large-Scale Image Recognition Without Normalization},
  journal={arXiv preprint arXiv:2102.06171},
  year={2021}
}

@inproceedings{brock2021characterizing,
  author={Andrew Brock and Soham De and Samuel L. Smith},
  title={Characterizing signal propagation to close the performance gap in
  unnormalized ResNets},
  booktitle={9th International Conference on Learning Representations, {ICLR}},
  year={2021}
}

@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}}
}