move deeplabv3 and resnext50 from model_zoo to model_zoo/official/cv

model_zoo/deeplabv3/README.md → model_zoo/official/cv/deeplabv3/README.md

@@ -1,11 +1,16 @@
-# Deeplab-V3 Example
+# DeeplabV3 Example
 
 ## Description
-This is an example of training DeepLabv3 with PASCAL VOC 2012 dataset in MindSpore. 
+This is an example of training DeepLabV3 with PASCAL VOC 2012 dataset in MindSpore. 
 
 ## Requirements
 - Install [MindSpore](https://www.mindspore.cn/install/en).
 - Download the VOC 2012 dataset for training.
+- We need to run `./src/remove_gt_colormap.py` to remove the label colormap. 
+  ``` bash
+  python remove_gt_colormap.py --original_gt_folder GT_FOLDER --output_dir OUTPUT_DIR
+
+  ```
 
 > Notes: 
   If you are running a fine-tuning or evaluation task, prepare the corresponding checkpoint file.
@@ -30,7 +35,7 @@ Set options in evaluation_config.py. Make sure the 'data_file' and 'finetune_ckp
 	```
 
 ## Options and Parameters
-It contains of parameters of Deeplab-V3 model and options for training, which is set in file config.py.
+It contains of parameters of DeeplabV3 model and options for training, which is set in file config.py.
 
 ### Options:
 ```

model_zoo/official/cv/deeplabv3/src/remove_gt_colormap.py

@@ -0,0 +1,76 @@
+# Copyright 2020 The Huawei Authors All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# ==============================================================================
+
+"""Removes the color map from segmentation annotations.
+Removes the color map from the ground truth segmentation annotations and save
+the results to output_dir.
+"""
+import glob
+import argparse
+import os.path
+import numpy as np
+
+
+from PIL import Image
+
+
+def _remove_colormap(filename):
+    """Removes the color map from the annotation.
+    Args:
+      filename: Ground truth annotation filename.
+    Returns:
+      Annotation without color map.
+    """
+    return np.array(Image.open(filename))
+
+
+def _save_annotation(annotation, filename):
+    """Saves the annotation as png file.
+    Args:
+      annotation: Segmentation annotation.
+      filename: Output filename.
+    """
+    pil_image = Image.fromarray(annotation.astype(dtype=np.uint8))
+    pil_image.save(filename, 'PNG')
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Demo of argparse")
+    parser.add_argument('--original_gt_folder', type=str, default='./VOCdevkit/VOC2012/SegmentationClass',
+                        help='Original ground truth annotations.')
+    parser.add_argument('--segmentation_format', type=str, default='png',
+                        help='Segmentation format.')
+    parser.add_argument('--output_dir', type=str, default='./VOCdevkit/VOC2012/SegmentationClassRaw',
+                        help='folder to save modified ground truth annotations.')
+    args = parser.parse_args()
+
+    # Create the output directory if not exists.
+
+    if not os.path.isdir(args.output_dir):
+        os.mkdir(args.output_dir)
+
+    annotations = glob.glob(os.path.join(args.original_gt_folder,
+                                         '*.' + args.segmentation_format))
+
+    for annotation in annotations:
+        raw_annotation = _remove_colormap(annotation)
+        filename = os.path.basename(annotation)[:-4]
+        _save_annotation(raw_annotation,
+                         os.path.join(
+                             args.output_dir,
+                             filename + '.' + args.segmentation_format))
+
+if __name__ == '__main__':
+    main()

model_zoo/resnext50/README.md → model_zoo/official/cv/resnext50/README.md

@@ -2,12 +2,12 @@
 
 ## Description
 
-This is an example of training ResNext50 with ImageNet dataset in Mindspore.
+This is an example of training ResNext50 in MindSpore.
 
 ## Requirements
 
 - Install [Mindspore](http://www.mindspore.cn/install/en).
-- Downlaod the dataset ImageNet2012.
+- Downlaod the dataset.
 
 ## Structure
 
@@ -91,9 +91,9 @@ sh run_standalone_train.sh DEVICE_ID DATA_PATH
 
 ```bash
 # distributed training example(8p)
-sh scripts/run_distribute_train.sh MINDSPORE_HCCL_CONFIG_PATH /ImageNet/train
+sh scripts/run_distribute_train.sh MINDSPORE_HCCL_CONFIG_PATH /dataset/train
 # standalone training example
-sh scripts/run_standalone_train.sh 0 /ImageNet_Original/train
+sh scripts/run_standalone_train.sh 0 /dataset/train
 ```
 
 #### Result
@@ -123,6 +123,6 @@ sh scripts/run_eval.sh 0 /opt/npu/datasets/classification/val /resnext50_100.ckp
 Evaluation result will be stored in the scripts path. Under this, you can find result like the followings in log.
  
 ```
-acc=78,16%(TOP1)
+acc=78.16%(TOP1)
 acc=93.88%(TOP5)
 ```