From 095eb3702fba3d8ebf3d44b7677a140aebae83fe Mon Sep 17 00:00:00 2001 From: liaogang Date: Mon, 6 Mar 2017 11:39:28 +0800 Subject: [PATCH] Fix conflict --- image_classification/{README_api.md => README.md} | 0 image_classification/deprecated/README.md | 8 +++----- 2 files changed, 3 insertions(+), 5 deletions(-) rename image_classification/{README_api.md => README.md} (100%) diff --git a/image_classification/README_api.md b/image_classification/README.md similarity index 100% rename from image_classification/README_api.md rename to image_classification/README.md diff --git a/image_classification/deprecated/README.md b/image_classification/deprecated/README.md index 05188be..f70b819 100644 --- a/image_classification/deprecated/README.md +++ b/image_classification/deprecated/README.md @@ -138,9 +138,9 @@ ResNet(Residual Network) \[[15](#参考文献)\] 是2015年ImageNet图像分类 ### 数据介绍与下载 -通用图像分类公开的标准数据集常用的有[CIFAR](https://www.cs.toronto.edu/~kriz/cifar.html)、[ImageNet](http://image-net.org/)、[COCO](http://mscoco.org/)等，常用的细粒度图像分类数据集包括[CUB-200-2011](http://www.vision.caltech.edu/visipedia/CUB-200-2011.html)、[Stanford Dog](http://vision.stanford.edu/aditya86/ImageNetDogs/)、[Oxford-flowers](http://www.robots.ox.ac.uk/~vgg/data/flowers/)等。其中ImageNet数据集规模相对较大，如[模型概览](#模型概览)一章所讲，大量研究成果基于ImageNet。ImageNet数据从2010年来稍有变化，常用的是ImageNet-2012数据集，该数据集包含1000个类别：训练集包含1,281,167张图片，每个类别数据732至1300张不等，验证集包含50,000张图片，平均每个类别50张图片。 +通用图像分类公开的标准数据集常用的有[CIFAR]()数据集。CIFAR10数据集包含60,000张32x32的彩色图片，10个类别，每个类包含6,000张。其中50,000张图片作为训练集，10000张作为测试集。图11从每个类别中随机抽取了10张图片，展示了所有的类别。


@@ -177,7 +177,7 @@ def initializer(settings, mean_path, is_train, **kwargs): } -@provider(init_hook=initializer, pool_size=50000) +@provider(init_hook=initializer, cache=CacheType.CACHE_PASS_IN_MEM) def process(settings, file_list): with open(file_list, 'r') as fdata: for fname in fdata: @@ -188,9 +188,7 @@ def process(settings, file_list): labels = batch['labels'] for im, lab in zip(images, labels): if settings.is_train and np.random.randint(2): - im = im.reshape(3, 32, 32) im = im[:,:,::-1] - im = im.flatten() im = im - settings.mean yield { 'image': im.astype('float32'), -- GitLab