文件列表:

LICENSE (1065, 2018-07-30)
imagenet.py (13326, 2018-07-30)
models (0, 2018-07-30)
models\__init__.py (0, 2018-07-30)
models\imagenet (0, 2018-07-30)
models\imagenet\__init__.py (290, 2018-07-30)
models\imagenet\densenet.py (6862, 2018-07-30)
models\imagenet\densenet_ibn_a.py (7784, 2018-07-30)
models\imagenet\resnet.py (6571, 2018-07-30)
models\imagenet\resnet_ibn_a.py (6639, 2018-07-30)
models\imagenet\resnet_ibn_a_old.py (7675, 2018-07-30)
models\imagenet\resnet_ibn_b.py (6268, 2018-07-30)
models\imagenet\resnext.py (5698, 2018-07-30)
models\imagenet\resnext_ibn_a.py (6628, 2018-07-30)
models\imagenet\se_module.py (610, 2018-07-30)
models\imagenet\se_resnet.py (3855, 2018-07-30)
models\imagenet\se_resnet_ibn_a.py (6380, 2018-07-30)
pretrained (0, 2018-07-30)
test.sh (300, 2018-07-30)
train.sh (361, 2018-07-30)
utils (0, 2018-07-30)
utils\IBNNet.png (75692, 2018-07-30)
utils\__init__.py (245, 2018-07-30)
utils\eval.py (523, 2018-07-30)
utils\images (0, 2018-07-30)
utils\images\cifar.png (344825, 2018-07-30)
utils\images\imagenet.png (46471, 2018-07-30)
utils\logger.py (4398, 2018-07-30)
utils\misc.py (2220, 2018-07-30)
utils\progress (0, 2018-07-30)
utils\progress\LICENSE (776, 2018-07-30)
utils\progress\MANIFEST.in (27, 2018-07-30)
utils\progress\demo.gif (924334, 2018-07-30)
utils\progress\progress (0, 2018-07-30)
utils\progress\progress\__init__.py (3188, 2018-07-30)
utils\progress\progress\bar.py (2839, 2018-07-30)
... ...

## Instance-Batch Normalization Network ### Paper Xingang Pan, Ping Luo, Jianping Shi, Xiaoou Tang. ["Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net"](https://arxiv.org/abs/1807.09441), ECCV2018. ### Introduction - IBN-Net carefully unifies instance normalization and batch normalization in a single deep network. - It provides an extremely simple way to increase both modeling and generalization capacity without adding model complexity. ### Requirements - Pytorch 0.3.1 (master branch) or Pytorch 0.4.1 (0.4.1 branch) ### Results Top1/Top5 error on the ImageNet validation set are reported. You may get different results when training your models with different random seed. | Model | origin | re-implementation | IBN-Net | | ------------------- | ------------------ | ------------------ | ------------------ | | DenseNet-121 | 25.0/- | 24.96/7.85 | 24.47/7.25 | | DenseNet-169 | 23.6/- | 24.02/7.06 | 23.25/6.51 | | ResNet-50 | 24.7/7.8 | 24.27/7.08 | 22.54/6.32 | | ResNet-101 | 23.6/7.1 | 22.48/6.23 | 21.39/5.59 | | ResNeXt-101 | 21.2/5.6 | 21.31/5.74 | 20.88/5.42 | | SE-ResNet-101 | 22.38/6.07 | 21.68/5.88 | 21.25/5.51 | ### Before Start 1. Clone the repository ```Shell git clone https://github.com/XingangPan/IBN-Net.git ``` 2. Download [ImageNet](http://image-net.org/download-images) dataset (if you need to test or train on ImageNet). You may follow the instruction at [fb.resnet.torch](https://github.com/facebook/fb.resnet.torch) to process the validation set. ### Testing 1. Download our pre-trained models and save them to `./pretrained`. Download link: [Pretrained models for pytorch0.3.1](https://drive.google.com/open?id=1JxSo6unmvwkCavEqh42NDKYUG29HoLE0), [Pretrained models for pytorch0.4.1](https://drive.google.com/open?id=1thS2B8UOSBi_cJX6zRy6YYRwz_nVFI_S) 2. Edit `test.sh`. Modify `model` and `data_path` to yours. Options for `model`: densenet121_ibn_a, densenet169_ibn_a, resnet50_ibn_a_old, resnet50_ibn_a, resnet50_ibn_b, resnet101_ibn_a_old, resnet101_ibn_a, resnext101_ibn_a, se_resnet101_ibn_a. (Note: For IBN-Net version of ResNet-50 and ResNet-101, our results in the paper are reported based on an slower implementation, corresponding to resnet50_ibn_a_old and resnet101_ibn_a_old here. We also provide a faster implementation, and the models are resnet50_ibn_a, resnet101_ibn_a, and all the rest. The top1/top5 error for resnet50_ibn_a and resnet101_ibn_a are 22.76/***1 and 21.29/5.61 respectively.) 3. Run test script ```Shell sh test.sh ``` ### Training 1. Edit `train.sh`. Modify `model` and `data_path` to yours. 2. Run train script ```Shell sh train.sh ``` This code is modified from [bearpaw/pytorch-classification](https://github.com/bearpaw/pytorch-classification). ### MXNet Implementation https://github.com/bruinxiong/IBN-Net.mxnet ### Citing IBN-Net ``` @inproceedings{pan2018IBN-Net, author = {Xingang Pan, Ping Luo, Jianping Shi, and Xiaoou Tang}, title = {Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net}, booktitle = {ECCV}, year = {2018} } ```

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：