|
|
# Res2Net
|
|
|
|
|
|
**Res2Net** is an image model that employs a variation on bottleneck residual blocks, [Res2Net Blocks](https://paperswithcode.com/method/res2net-block). The motivation is to be able to represent features at multiple scales. This is achieved through a novel building block for CNNs that constructs hierarchical residual-like connections within one single residual block. This represents multi-scale features at a granular level and increases the range of receptive fields for each network layer.
|
|
|
|
|
|
{% 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
|
|
|
@article{Gao_2021,
|
|
|
title={Res2Net: A New Multi-Scale Backbone Architecture},
|
|
|
volume={43},
|
|
|
ISSN={1939-3539},
|
|
|
url={http://dx.doi.org/10.1109/TPAMI.2019.2938758},
|
|
|
DOI={10.1109/tpami.2019.2938758},
|
|
|
number={2},
|
|
|
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
|
|
|
publisher={Institute of Electrical and Electronics Engineers (IEEE)},
|
|
|
author={Gao, Shang-Hua and Cheng, Ming-Ming and Zhao, Kai and Zhang, Xin-Yu and Yang, Ming-Hsuan and Torr, Philip},
|
|
|
year={2021},
|
|
|
month={Feb},
|
|
|
pages={652–662}
|
|
|
}
|
|
|
```
|
|
|
|
|
|
<!--
|
|
|
Models:
|
|
|
- Name: res2net101_26w_4s
|
|
|
Metadata:
|
|
|
FLOPs: 10415881200
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 181456059
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net101_26w_4s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L152
|
|
|
In Collection: Res2Net
|
|
|
- Name: res2net50_26w_6s
|
|
|
Metadata:
|
|
|
FLOPs: 8130156528
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 148603239
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net50_26w_6s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L163
|
|
|
In Collection: Res2Net
|
|
|
- Name: res2net50_26w_8s
|
|
|
Metadata:
|
|
|
FLOPs: 10760338992
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 194085165
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net50_26w_8s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L174
|
|
|
In Collection: Res2Net
|
|
|
- Name: res2net50_14w_8s
|
|
|
Metadata:
|
|
|
FLOPs: 5403546768
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 100638543
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net50_14w_8s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L196
|
|
|
In Collection: Res2Net
|
|
|
- Name: res2net50_26w_4s
|
|
|
Metadata:
|
|
|
FLOPs: 5499974064
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 103110087
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net50_26w_4s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L141
|
|
|
In Collection: Res2Net
|
|
|
- Name: res2net50_48w_2s
|
|
|
Metadata:
|
|
|
FLOPs: 5375291520
|
|
|
Epochs: 100
|
|
|
Batch Size: 256
|
|
|
Training Data:
|
|
|
- ImageNet
|
|
|
Training Techniques:
|
|
|
- SGD with Momentum
|
|
|
- Weight Decay
|
|
|
Training Resources: 4x Titan Xp GPUs
|
|
|
Architecture:
|
|
|
- Batch Normalization
|
|
|
- Convolution
|
|
|
- Global Average Pooling
|
|
|
- ReLU
|
|
|
- Res2Net Block
|
|
|
File Size: 101421406
|
|
|
Tasks:
|
|
|
- Image Classification
|
|
|
ID: res2net50_48w_2s
|
|
|
LR: 0.1
|
|
|
Crop Pct: '0.875'
|
|
|
Momentum: 0.9
|
|
|
Image Size: '224'
|
|
|
Weight Decay: 0.0001
|
|
|
Interpolation: bilinear
|
|
|
Code: https://github.com/rwightman/pytorch-image-models/blob/d8e69206be253892b2956341fea09fdebfaae4e3/timm/models/res2net.py#L185
|
|
|
In Collection: Res2Net
|
|
|
Collections:
|
|
|
- Name: Res2Net
|
|
|
Paper:
|
|
|
title: 'Res2Net: A New Multi-scale Backbone Architecture'
|
|
|
url: https://paperswithcode.com//paper/res2net-a-new-multi-scale-backbone
|
|
|
type: model-index
|
|
|
Type: model-index
|
|
|
-->
|