new branch dev

README.md

@@ -1,4 +1,4 @@
-# refinenet（in work condition）
+# refinenet
 a tensorflow implement of refinenet. RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation
 
 
@@ -19,7 +19,7 @@ this is a tensorflow implement of refinenet discribed in [arxiv:1611.06612](http
 - put images in demo/ and run python RefineNet/demo.py 
 
 ## roadmap
-- [ ] python2/3 compatibility
+- [x] python2/3 compatibility
 - [ ] Complete realization of refinenet model
 - [ ] test on pascal voc, give the IoU result
 - [ ] training on other datasets

RefineNet/demo.py

@@ -2,11 +2,13 @@
 import cv2
 import time
 import os,shutil
+import sys
 import numpy as np
 import tensorflow as tf
 slim = tf.contrib.slim
 
-import model as model
+sys.path.append(os.getcwd())
+from nets import model as model
 from matplotlib import pyplot as plt
 from utils.pascal_voc import pascal_segmentation_lut
 from utils.visualization import visualize_segmentation_adaptive

RefineNet/multi_gpu_train.py

@@ -5,36 +5,35 @@
 import shutil
 import datetime
 import os
-import pickle
 import cv2
+import pickle
 import numpy as np
 from tensorflow.contrib import slim
 import sys
 sys.path.append(os.getcwd())
-import model as model
-from utils.input import get_batch,generator
+from nets import model as model
 from utils.tf_records import read_tfrecord_and_decode_into_image_annotation_pair_tensors
 from utils.pascal_voc import pascal_segmentation_lut
 from utils.augmentation import (distort_randomly_image_color,flip_randomly_left_right_image_with_annotation,
                                 scale_randomly_image_with_annotation_with_fixed_size_output)
 
 
-tf.app.flags.DEFINE_integer('batch_size', 2, '')
-tf.app.flags.DEFINE_integer('train_size', 384, '')
+tf.app.flags.DEFINE_integer('batch_size', 3, '')
+tf.app.flags.DEFINE_integer('train_size', 512, '')
 tf.app.flags.DEFINE_float('learning_rate', 0.0001, '')
-tf.app.flags.DEFINE_integer('max_steps', 100000, '')
+tf.app.flags.DEFINE_integer('max_steps', 60000, '')
 tf.app.flags.DEFINE_float('moving_average_decay', 0.997, '')
 tf.app.flags.DEFINE_integer('num_classes', 21, '')
-tf.app.flags.DEFINE_string('gpu_list', '0', '')
+tf.app.flags.DEFINE_string('gpu_list', '0,1', '')
 tf.app.flags.DEFINE_string('checkpoint_path', 'checkpoints/', '')
 tf.app.flags.DEFINE_string('logs_path', 'logs/', '')
-tf.app.flags.DEFINE_boolean('restore', False, 'whether to resotre from checkpoint')
+tf.app.flags.DEFINE_boolean('restore', True, 'whether to resotre from checkpoint')
 tf.app.flags.DEFINE_integer('save_checkpoint_steps', 2000, '')
 tf.app.flags.DEFINE_integer('save_summary_steps', 10, '')
-tf.app.flags.DEFINE_integer('save_image_steps', 10, '')
+tf.app.flags.DEFINE_integer('save_image_steps', 100, '')
 tf.app.flags.DEFINE_string('training_data_path', 'data/pascal_augmented_train.tfrecords', '')
 tf.app.flags.DEFINE_string('pretrained_model_path', 'data/resnet_v1_101.ckpt', '')
-tf.app.flags.DEFINE_integer('decay_steps',40000,'')
+tf.app.flags.DEFINE_integer('decay_steps',20000,'')
 tf.app.flags.DEFINE_integer('decay_rate',0.1,'')
 FLAGS = tf.app.flags.FLAGS
 
@@ -94,10 +93,6 @@ def main(argv=None):
                 shutil.rmtree(FLAGS.checkpoint_path)
                 os.makedirs(FLAGS.checkpoint_path)
 
-
-    #input_images = tf.placeholder(tf.float32, shape=[None, None, None, 3], name='input_images')
-    #input_segs = tf.placeholder(tf.float32, shape=[None, None,None, 1], name='input_segs')
-
     filename_queue = tf.train.string_input_producer([FLAGS.training_data_path], num_epochs=1000)
     image, annotation = read_tfrecord_and_decode_into_image_annotation_pair_tensors(filename_queue)
 
@@ -177,17 +172,12 @@ def main(argv=None):
             if FLAGS.pretrained_model_path is not None:
                 variable_restore_op(sess)
 
-        #data_generator = get_batch(num_workers=8, batch_size=FLAGS.batch_size * len(gpus))
-        #data_generator = generator(batch_size=FLAGS.batch_size * len(gpus))
         start = time.time()
         coord = tf.train.Coordinator()
         threads = tf.train.start_queue_runners(coord=coord)
         try:
             while not coord.should_stop():
                 for step in range(restore_step,FLAGS.max_steps):
-
-                    #data = next(data_generator)
-
                     if step != 0 and step % FLAGS.decay_steps == 0:
                         sess.run(tf.assign(learning_rate, learning_rate.eval() * FLAGS.decay_rate))
 
@@ -219,6 +209,8 @@ def main(argv=None):
                         seg_split=np.squeeze(seg_split)[0]
                         pred=np.squeeze(pred)[0]
 
+                        #img_split=cv2.resize(img_split,(128,128))
+
                         color_seg = np.zeros((seg_split.shape[0], seg_split.shape[1], 3))
                         for i in range(seg_split.shape[0]):
                             for j in range(seg_split.shape[1]):
@@ -229,11 +221,10 @@ def main(argv=None):
                             for j in range(pred.shape[1]):
                                 color_pred[i, j, :] = color_map[str(pred[i][j])]
 
-                        write_img=np.hstack((img_split,color_seg,color_pred))
+                        write_img=np.hstack((color_seg,color_pred))
                         log_image_summary_op = sess.run(log_image,feed_dict={log_image_name: log_image_name_str, \
                                                                    log_image_data: write_img})
                         summary_writer.add_summary(log_image_summary_op, global_step=step)
-
         except tf.errors.OutOfRangeError:
             print('finish')
         finally: