You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
pytorch-image-models/hfdocs/source/models/skresnet.mdx

179 lines
5.5 KiB

# SK-ResNet
**SK ResNet** is a variant of a [ResNet](https://www.paperswithcode.com/method/resnet) that employs a [Selective Kernel](https://paperswithcode.com/method/selective-kernel) unit. In general, all the large kernel convolutions in the original bottleneck blocks in ResNet are replaced by the proposed [SK convolutions](https://paperswithcode.com/method/selective-kernel-convolution), enabling the network to choose appropriate receptive field sizes in an adaptive manner.
## How do I use this model on an image?
To load a pretrained model:
```py
>>> import timm
>>> model = timm.create_model('skresnet18', pretrained=True)
>>> model.eval()
```
To load and preprocess the image:
```py
>>> import urllib
>>> from PIL import Image
>>> from timm.data import resolve_data_config
>>> from timm.data.transforms_factory import create_transform
>>> config = resolve_data_config({}, model=model)
>>> transform = create_transform(**config)
>>> url, filename = ("https://github.com/pytorch/hub/raw/master/images/dog.jpg", "dog.jpg")
>>> urllib.request.urlretrieve(url, filename)
>>> img = Image.open(filename).convert('RGB')
>>> tensor = transform(img).unsqueeze(0) # transform and add batch dimension
```
To get the model predictions:
```py
>>> import torch
>>> with torch.no_grad():
... out = model(tensor)
>>> probabilities = torch.nn.functional.softmax(out[0], dim=0)
>>> print(probabilities.shape)
>>> # prints: torch.Size([1000])
```
To get the top-5 predictions class names:
```py
>>> # Get imagenet class mappings
>>> url, filename = ("https://raw.githubusercontent.com/pytorch/hub/master/imagenet_classes.txt", "imagenet_classes.txt")
>>> urllib.request.urlretrieve(url, filename)
>>> with open("imagenet_classes.txt", "r") as f:
... categories = [s.strip() for s in f.readlines()]
>>> # Print top categories per image
>>> top5_prob, top5_catid = torch.topk(probabilities, 5)
>>> for i in range(top5_prob.size(0)):
... print(categories[top5_catid[i]], top5_prob[i].item())
>>> # prints class names and probabilities like:
>>> # [('Samoyed', 0.6425196528434753), ('Pomeranian', 0.04062102362513542), ('keeshond', 0.03186424449086189), ('white wolf', 0.01739676296710968), ('Eskimo dog', 0.011717947199940681)]
```
Replace the model name with the variant you want to use, e.g. `skresnet18`. You can find the IDs in the model summaries at the top of this page.
To extract image features with this model, follow the [timm feature extraction examples](../feature_extraction), just change the name of the model you want to use.
## How do I finetune this model?
You can finetune any of the pre-trained models just by changing the classifier (the last layer).
```py
>>> model = timm.create_model('skresnet18', pretrained=True, num_classes=NUM_FINETUNE_CLASSES)
```
To finetune on your own dataset, you have to write a training loop or adapt [timm's training
script](https://github.com/rwightman/pytorch-image-models/blob/master/train.py) to use your dataset.
## How do I train this model?
You can follow the [timm recipe scripts](../scripts) for training a new model afresh.
## Citation
```BibTeX
@misc{li2019selective,
title={Selective Kernel Networks},
author={Xiang Li and Wenhai Wang and Xiaolin Hu and Jian Yang},
year={2019},
eprint={1903.06586},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
<!--
Type: model-index
Collections:
- Name: SKResNet
Paper:
Title: Selective Kernel Networks
URL: https://paperswithcode.com/paper/selective-kernel-networks
Models:
- Name: skresnet18
In Collection: SKResNet
Metadata:
FLOPs: 2333467136
Parameters: 11960000
File Size: 47923238
Architecture:
- Convolution
- Dense Connections
- Global Average Pooling
- Max Pooling
- Residual Connection
- Selective Kernel
- Softmax
Tasks:
- Image Classification
Training Techniques:
- SGD with Momentum
- Weight Decay
Training Data:
- ImageNet
Training Resources: 8x GPUs
ID: skresnet18
LR: 0.1
Epochs: 100
Layers: 18
Crop Pct: '0.875'
Momentum: 0.9
Batch Size: 256
Image Size: '224'
Weight Decay: 4.0e-05
Interpolation: bicubic
Code: https://github.com/rwightman/pytorch-image-models/blob/a7f95818e44b281137503bcf4b3e3e94d8ffa52f/timm/models/sknet.py#L148
Weights: https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/skresnet18_ra-4eec2804.pth
Results:
- Task: Image Classification
Dataset: ImageNet
Metrics:
Top 1 Accuracy: 73.03%
Top 5 Accuracy: 91.17%
- Name: skresnet34
In Collection: SKResNet
Metadata:
FLOPs: 4711849952
Parameters: 22280000
File Size: 89299314
Architecture:
- Convolution
- Dense Connections
- Global Average Pooling
- Max Pooling
- Residual Connection
- Selective Kernel
- Softmax
Tasks:
- Image Classification
Training Techniques:
- SGD with Momentum
- Weight Decay
Training Data:
- ImageNet
Training Resources: 8x GPUs
ID: skresnet34
LR: 0.1
Epochs: 100
Layers: 34
Crop Pct: '0.875'
Momentum: 0.9
Batch Size: 256
Image Size: '224'
Weight Decay: 4.0e-05
Interpolation: bicubic
Code: https://github.com/rwightman/pytorch-image-models/blob/a7f95818e44b281137503bcf4b3e3e94d8ffa52f/timm/models/sknet.py#L165
Weights: https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/skresnet34_ra-bdc0ccde.pth
Results:
- Task: Image Classification
Dataset: ImageNet
Metrics:
Top 1 Accuracy: 76.93%
Top 5 Accuracy: 93.32%
-->