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.
108 lines
2.7 KiB
108 lines
2.7 KiB
# SE ResNet
|
|
|
|
**SE ResNet** is a variant of a [ResNet](https://www.paperswithcode.com/method/resnet) that employs [squeeze-and-excitation blocks](https://paperswithcode.com/method/squeeze-and-excitation-block) to enable the network to perform dynamic channel-wise feature recalibration.
|
|
|
|
{% include 'code_snippets.md' %}
|
|
|
|
## How do I train this model?
|
|
|
|
You can follow the [timm recipe scripts](https://rwightman.github.io/pytorch-image-models/scripts/) for training a new model afresh.
|
|
|
|
## Citation
|
|
|
|
```BibTeX
|
|
@misc{hu2019squeezeandexcitation,
|
|
title={Squeeze-and-Excitation Networks},
|
|
author={Jie Hu and Li Shen and Samuel Albanie and Gang Sun and Enhua Wu},
|
|
year={2019},
|
|
eprint={1709.01507},
|
|
archivePrefix={arXiv},
|
|
primaryClass={cs.CV}
|
|
}
|
|
```
|
|
|
|
<!--
|
|
Models:
|
|
- Name: seresnet152d
|
|
Metadata:
|
|
FLOPs: 20161904304
|
|
Epochs: 100
|
|
Batch Size: 1024
|
|
Training Data:
|
|
- ImageNet
|
|
Training Techniques:
|
|
- Label Smoothing
|
|
- SGD with Momentum
|
|
- Weight Decay
|
|
Training Resources: 8x NVIDIA Titan X GPUs
|
|
Architecture:
|
|
- 1x1 Convolution
|
|
- Batch Normalization
|
|
- Bottleneck Residual Block
|
|
- Convolution
|
|
- Global Average Pooling
|
|
- Max Pooling
|
|
- ReLU
|
|
- Residual Block
|
|
- Residual Connection
|
|
- Softmax
|
|
- Squeeze-and-Excitation Block
|
|
File Size: 268144497
|
|
Tasks:
|
|
- Image Classification
|
|
ID: seresnet152d
|
|
LR: 0.6
|
|
Layers: 152
|
|
Dropout: 0.2
|
|
Crop Pct: '0.94'
|
|
Momentum: 0.9
|
|
Image Size: '256'
|
|
Interpolation: bicubic
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/a7f95818e44b281137503bcf4b3e3e94d8ffa52f/timm/models/resnet.py#L1206
|
|
In Collection: SE ResNet
|
|
- Name: seresnet50
|
|
Metadata:
|
|
FLOPs: 5285062320
|
|
Epochs: 100
|
|
Batch Size: 1024
|
|
Training Data:
|
|
- ImageNet
|
|
Training Techniques:
|
|
- Label Smoothing
|
|
- SGD with Momentum
|
|
- Weight Decay
|
|
Training Resources: 8x NVIDIA Titan X GPUs
|
|
Architecture:
|
|
- 1x1 Convolution
|
|
- Batch Normalization
|
|
- Bottleneck Residual Block
|
|
- Convolution
|
|
- Global Average Pooling
|
|
- Max Pooling
|
|
- ReLU
|
|
- Residual Block
|
|
- Residual Connection
|
|
- Softmax
|
|
- Squeeze-and-Excitation Block
|
|
File Size: 112621903
|
|
Tasks:
|
|
- Image Classification
|
|
ID: seresnet50
|
|
LR: 0.6
|
|
Layers: 50
|
|
Dropout: 0.2
|
|
Crop Pct: '0.875'
|
|
Momentum: 0.9
|
|
Image Size: '224'
|
|
Interpolation: bicubic
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/a7f95818e44b281137503bcf4b3e3e94d8ffa52f/timm/models/resnet.py#L1180
|
|
In Collection: SE ResNet
|
|
Collections:
|
|
- Name: SE ResNet
|
|
Paper:
|
|
title: Squeeze-and-Excitation Networks
|
|
url: https://paperswithcode.com//paper/squeeze-and-excitation-networks
|
|
type: model-index
|
|
Type: model-index
|
|
-->
|