|
# Download the pretrained checkpoints |
|
|
|
To facilitate the model training, we also provide some checkpoints that are |
|
pretrained on ImageNet. |
|
|
|
After downloading the desired pretrained checkpoint, remember to update |
|
the `initial_checkpoint` path in the config files. |
|
|
|
## Checkpoints |
|
|
|
**Simple Training Strategy**: This training strategy yields a similar |
|
performance to the original ResNet paper [2]. |
|
|
|
Backbone | Pretrained Dataset |
|
-------- | :---------------: |
|
ResNet-50 ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/resnet50_imagenet1k.tar.gz)) | ImageNet-1K |
|
|
|
**Strong Training Strategy**: This training strategy additionally |
|
employs AutoAugment [3], label-smoothing [4], and drop-path [5], yielding |
|
a stronger performance on ImageNet than the original ResNet paper [2]. |
|
|
|
Backbone | Pretrained Dataset |
|
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | :----------------: |
|
ResNet-50 ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/resnet50_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
ResNet-50-Beta ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/resnet50_beta_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
Wide-ResNet-41 ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/wide_resnet41_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
SWideRNet-SAC-(1, 1, 1) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/swidernet_sac_1_1_1_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
SWideRNet-SAC-(1, 1, 3) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/swidernet_sac_1_1_3_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
SWideRNet-SAC-(1, 1, 4.5) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/swidernet_sac_1_1_4.5_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
Axial-SWideRNet-(1, 1, 1) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/axial_swidernet_1_1_1_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
Axial-SWideRNet-(1, 1, 3) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/axial_swidernet_1_1_3_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
Axial-SWideRNet-(1, 1, 4.5) ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/axial_swidernet_1_1_4.5_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
MaX-DeepLab-S-Backbone ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/max_deeplab_s_backbone_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
MaX-DeepLab-L-Backbone ([initial_checkpoint](https://storage.googleapis.com/gresearch/tf-deeplab/checkpoint/max_deeplab_l_backbone_imagenet1k_strong_training_strategy.tar.gz)) | ImageNet-1K |
|
|
|
### References |
|
|
|
1. Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, |
|
Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, |
|
Michael Bernstein, Alexander C. Berg, and Li Fei-Fei. "ImageNet Large |
|
Scale Visual Recognition Challenge". IJCV, 2015. |
|
|
|
2. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. "Deep residual |
|
learning for image recognition. In CVPR, 2016. |
|
|
|
3. Ekin D Cubuk, Barret Zoph, Dandelion Mane, Vijay Vasudevan, and |
|
Quoc V Le. "Autoaugment: Learning augmentation policies from data". |
|
In CVPR, 2019. |
|
|
|
4. Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jonathon Shlens, and |
|
Zbigniew Wojna. "Rethinking the inception architecture for computer |
|
vision." In CVPR, 2016. |
|
|
|
5. Gao Huang, Yu Sun, Zhuang Liu, Daniel Sedra, and Kilian Q Weinberger. |
|
"Deep networks with stochastic depth." In ECCV, 2016. |
|
|