You can deploy this model and start earning money today!

BVRA/
legacy_seresnext101_32x4d.in1k_ft_df20m_224

Name: legacy_seresnext101_32x4d.in1k_ft_df20m_224
Availability: InStock

Views

No views yet

Model card for BVRA/legacy_seresnext101_32x4d.in1k_ft_df20m_224

Model Details

Model Type: Danish Fungi Classification
Model Stats:
- Params (M): 47.3M
- Image size: 224 x 224
Papers:
- Original 1: Aggregated Residual Transformations for Deep Neural Networks --> https://arxiv.org/pdf/1611.05431v2
- Original 2: Squeeze-and-Excitation Networks --> https://arxiv.org/pdf/1709.01507v4
- Train Dataset: DF20 --> https://github.com/BohemianVRA/DanishFungiDataset/

Model Usage

Image Embeddings

python

import timm
import torch
import torchvision.transforms as T
from PIL import Image
from urllib.request import urlopen
model = timm.create_model("hf-hub:BVRA/legacy_seresnext101_32x4d.in1k_ft_df20m_224", pretrained=True)
model = model.eval()
train_transforms = T.Compose([T.Resize((224, 224)), 
                              T.ToTensor(), 
                              T.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]) 
img = Image.open(PATH_TO_YOUR_IMAGE)
output = model(train_transforms(img).unsqueeze(0))  # output is (batch_size, num_features) shaped tensor

Citation

bibtex

@InProceedings{Picek_2022_WACV,
    author    = {Picek, Lukas and Sulc, Milan and Matas, Jiri and Jeppesen, Thomas S. and Heilmann-Clausen, Jacob and L{e}ss{\o}e, Thomas and Fr{\o}slev, Tobias},
    title     = {Danish Fungi 2020 - Not Just Another Image Recognition Dataset},
    booktitle = {Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
    month     = {January},
    year      = {2022},
    pages     = {1525-1535}
}

@article{picek2022automatic,
  title={Automatic Fungi Recognition: Deep Learning Meets Mycology},
  author={Picek, Lukas and Sulc, Milan and Matas, Jiri and Heilmann-Clausen, Jacob and Jeppesen, Thomas S and Lind, Emil},
  journal={Sensors},
  volume={22},
  number={2},
  pages={633},
  year={2022},
  publisher={Multidisciplinary Digital Publishing Institute}
}