Merge pull request #1552 from shippingwang/add_image_classification_pretrained_models

add image classification new pretrained models

fluid/PaddleCV/image_classification/README.md

@@ -196,10 +196,19 @@ Models are trained by starting with learning rate ```0.1``` and decaying it by `
 |model | top-1/top-5 accuracy(PIL)| top-1/top-5 accuracy(CV2) |
 |- |:-: |:-:|
 |[AlexNet](http://paddle-imagenet-models-name.bj.bcebos.com/AlexNet_pretrained.zip) | 56.71%/79.18% | 55.88%/78.65% |
-|[VGG11](http://paddle-imagenet-models-name.bj.bcebos.com/VGG11_pretained.zip) | 68.92%/88.66% | 68.61%/88.60% |
+|[VGG11](https://paddle-imagenet-models-name.bj.bcebos.com/VGG11_pretrained.zip) | 69.22%/89.09% | 69.01%/88.90% |
+|[VGG13](https://paddle-imagenet-models-name.bj.bcebos.com/VGG13_pretrained.zip) | 70.14%/89.48% | 69.83%/89.13% |
+|[VGG16](https://paddle-imagenet-models-name.bj.bcebos.com/VGG16_pretrained.zip) | 72.08%/90.63% | 71.65%/90.57% |
+|[VGG19](https://paddle-imagenet-models-name.bj.bcebos.com/VGG19_pretrained.zip) | 72.56%/90.83% | 72.32%/90.98% |
 |[MobileNetV1](http://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV1_pretrained.zip) | 70.91%/89.54% | 70.51%/89.35% |
 |[ResNet50](http://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_pretrained.zip) | 76.35%/92.80% | 76.22%/92.92% |
 |[ResNet101](http://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.zip) | 77.49%/93.57% | 77.56%/93.64% |
+|[ResNet152](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet152_pretrained.zip) | 78.12%/93.93% | 77.92%/93.87% |
+|[SE_ResNeXt50_32x4d](https://paddle-imagenet-models-name.bj.bcebos.com/SE_ResNext50_32x4d_pretrained.zip) | 78.50%/94.01% | 78.44%/93.96% |
+|[SE_ResNeXt101_32x4d](https://paddle-imagenet-models-name.bj.bcebos.com/SE_ResNeXt101_32x4d_pretrained.zip) | 79.26%/94.22% | 79.12%/94.20% |
+
+
+
 
 - Released models: not specify parameter names
 

fluid/PaddleCV/image_classification/README_cn.md

@@ -194,10 +194,16 @@ Models包括两种模型：带有参数名字的模型，和不带有参数名
 |model | top-1/top-5 accuracy(PIL)| top-1/top-5 accuracy(CV2) |
 |- |:-: |:-:|
 |[AlexNet](http://paddle-imagenet-models-name.bj.bcebos.com/AlexNet_pretrained.zip) | 56.71%/79.18% | 55.88%/78.65% |
-|[VGG11](http://paddle-imagenet-models-name.bj.bcebos.com/VGG11_pretained.zip) | 68.92%/88.66% | 68.61%/88.60% |
+|[VGG11](https://paddle-imagenet-models-name.bj.bcebos.com/VGG11_pretrained.zip) | 69.22%/89.09% | 69.01%/88.90% |
+|[VGG13](https://paddle-imagenet-models-name.bj.bcebos.com/VGG13_pretrained.zip) | 70.14%/89.48% | 69.83%/89.13% |
+|[VGG16](https://paddle-imagenet-models-name.bj.bcebos.com/VGG16_pretrained.zip) | 72.08%/90.63% | 71.65%/90.57% |
+|[VGG19](https://paddle-imagenet-models-name.bj.bcebos.com/VGG19_pretrained.zip) | 72.56%/90.83% | 72.32%/90.98% |
 |[MobileNetV1](http://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV1_pretrained.zip) | 70.91%/89.54% | 70.51%/89.35% |
 |[ResNet50](http://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_pretrained.zip) | 76.35%/92.80% | 76.22%/92.92% |
 |[ResNet101](http://paddle-imagenet-models-name.bj.bcebos.com/ResNet101_pretrained.zip) | 77.49%/93.57% | 77.56%/93.64% |
+|[ResNet152](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet152_pretrained.zip) | 78.12%/93.93% | 77.92%/93.87% |
+|[SE_ResNeXt50_32x4d](https://paddle-imagenet-models-name.bj.bcebos.com/SE_ResNext50_32x4d_pretrained.zip) | 78.50%/94.01% | 78.44%/93.96% |
+|[SE_ResNeXt101_32x4d](https://paddle-imagenet-models-name.bj.bcebos.com/SE_ResNeXt101_32x4d_pretrained.zip) | 79.26%/94.22% | 79.12%/94.20% |
 
 - Released models: not specify parameter names
 

fluid/PaddleCV/image_classification/eval.py

@@ -7,7 +7,8 @@ import time
 import sys
 import paddle
 import paddle.fluid as fluid
-import models
+#import models
+import models_name as models
 #import reader_cv2 as reader
 import reader as reader
 import argparse
@@ -84,7 +85,7 @@ def eval(args):
 
         fluid.io.load_vars(exe, pretrained_model, predicate=if_exist)
 
-    val_reader = paddle.batch(reader.val(""), batch_size=args.batch_size)
+    val_reader = paddle.batch(reader.val(), batch_size=args.batch_size)
     feeder = fluid.DataFeeder(place=place, feed_list=[image, label])
 
     fetch_list = [avg_cost.name, acc_top1.name, acc_top5.name]

fluid/PaddleCV/image_classification/reader_cv2.py

@@ -101,8 +101,6 @@ def process_image(sample,
     std = [0.229, 0.224, 0.225] if std is None else std
 
     img_path = sample[0]
-    print('&' * 80)
-    print(img_path)
     img = cv2.imread(img_path)
 
     if mode == 'train':

fluid/PaddleCV/image_classification/run.sh

@@ -78,3 +78,58 @@ python train.py \
 #	--num_epochs=120 \
 #       --lr=0.1
 
+#ResNet152:
+#python train.py \
+#       --model=ResNet152 \
+#       --batch_size=256 \
+#       --total_images=1281167 \
+#       --image_shape=3,224,224 \
+#       --lr_strategy=piecewise_decay \
+#       --lr=0.1 \
+#       --num_epochs=120 \
+#       --l2_decay=1e-4 \(TODO)
+
+
+#SE_ResNeXt50:
+#python train.py \
+#       --model=SE_ResNeXt50 \
+#       --batch_size=400 \
+#       --total_images=1281167 \
+#       --image_shape=3,224,224 \
+#       --lr_strategy=cosine_decay \
+#       --lr=0.1 \
+#       --num_epochs=200 \
+#       --l2_decay=12e-5 \(TODO)
+
+#SE_ResNeXt101:
+#python train.py \
+#        --model=SE_ResNeXt101 \
+#        --batch_size=400 \
+#        --total_images=1281167 \
+#        --image_shape=3,224,224 \
+#        --lr_strategy=cosine_decay \
+#        --lr=0.1 \
+#        --num_epochs=200 \
+#        --l2_decay=15e-5 \(TODO)
+
+#VGG11:
+#python train.py \
+#        --model=VGG11 \
+#        --batch_size=512 \
+#        --total_images=1281167 \
+#        --image_shape=3,224,224 \
+#        --lr_strategy=cosine_decay \
+#        --lr=0.1 \
+#        --num_epochs=90 \
+#        --l2_decay=2e-4 \(TODO)
+
+#VGG13:
+#python train.py
+#        --model=VGG13 \          
+#        --batch_size=256 \
+#        --total_images=1281167 \
+#        --image_shape=3,224,224 \
+#        --lr_strategy=cosine_decay \
+#        --lr=0.01 \
+#        --num_epochs=90 \
+#        --l2_decay=3e-4 \(TODO)