From fa1bf0b60ebc5d0e49c37f4cb0edf27a3bf5ab8f Mon Sep 17 00:00:00 2001
From: tomrink <rink@ssec.wisc.edu>
Date: Sun, 7 Aug 2022 11:19:34 -0500
Subject: [PATCH] snapshot...
---
modules/deeplearning/espcn.py | 41 +++++++++++++++++++++++++++--------
1 file changed, 32 insertions(+), 9 deletions(-)
diff --git a/modules/deeplearning/espcn.py b/modules/deeplearning/espcn.py
index 2240ad97..b7e561c9 100644
--- a/modules/deeplearning/espcn.py
+++ b/modules/deeplearning/espcn.py
@@ -23,10 +23,10 @@ else:
NumLogits = NumClasses
BATCH_SIZE = 64
-NUM_EPOCHS = 60
+NUM_EPOCHS = 80
TRACK_MOVING_AVERAGE = False
-EARLY_STOP = False
+EARLY_STOP = True
NOISE_TRAINING = False
NOISE_STDDEV = 0.10
@@ -366,7 +366,7 @@ class ESPCN:
self.get_evaluate_dataset(idxs)
- def build_espcn(self):
+ def build_espcn(self, do_drop_out=False, do_batch_norm=False, drop_rate=0.5):
print('build_cnn')
# padding = "VALID"
padding = "SAME"
@@ -389,12 +389,20 @@ class ESPCN:
if NOISE_TRAINING:
conv = tf.keras.layers.GaussianNoise(stddev=NOISE_STDDEV)(conv)
+ if do_drop_out:
+ conv = tf.keras.layers.Dropout(drop_rate)(conv)
+ if do_batch_norm:
+ conv = tf.keras.layers.BatchNormalization()(conv)
+
conv = tf.keras.layers.Conv2D(num_filters, kernel_size=5, strides=1, padding=padding, activation=activation)(conv)
- conv = tf.keras.layers.BatchNormalization()(conv)
print(conv.shape)
+ if do_drop_out:
+ conv = tf.keras.layers.Dropout(drop_rate)(conv)
+ if do_batch_norm:
+ conv = tf.keras.layers.BatchNormalization()(conv)
+
conv = tf.keras.layers.Conv2D(num_filters, kernel_size=3, strides=1, padding=padding, activation=activation)(conv)
- conv = tf.keras.layers.BatchNormalization()(conv)
print(conv.shape)
# conv = tf.keras.layers.Conv2D(num_filters, kernel_size=3, strides=1, padding=padding, activation=None)(conv)
@@ -405,12 +413,21 @@ class ESPCN:
# conv = tf.keras.layers.LeakyReLU()(conv)
# print(conv.shape)
+ if do_drop_out:
+ conv = tf.keras.layers.Dropout(drop_rate)(conv)
+ if do_batch_norm:
+ conv = tf.keras.layers.BatchNormalization()(conv)
+
conv = tf.keras.layers.Conv2D(num_filters // 2, kernel_size=3, strides=1, padding=padding, activation=activation)(conv)
conv = tf.keras.layers.BatchNormalization()(conv)
print(conv.shape)
+ if do_drop_out:
+ conv = tf.keras.layers.Dropout(drop_rate)(conv)
+ if do_batch_norm:
+ conv = tf.keras.layers.BatchNormalization()(conv)
+
conv = tf.keras.layers.Conv2D(num_filters // 2, kernel_size=3, strides=1, padding=padding, activation=activation)(conv)
- conv = tf.keras.layers.BatchNormalization()(conv)
print(conv.shape)
# conv = tf.keras.layers.Conv2D(4, kernel_size=3, strides=1, padding=padding, activation=activation)(conv)
@@ -419,7 +436,11 @@ class ESPCN:
conv = tf.keras.layers.Conv2DTranspose(num_filters // 4, kernel_size=3, strides=2, padding=padding, activation=activation)(conv)
print(conv.shape)
- self.logits = tf.keras.layers.Conv2D(1, kernel_size=1, strides=1, padding=padding, name='probability', activation=tf.nn.sigmoid)(conv)
+ conv = tf.keras.layers.Conv2DTranspose(num_filters // 4, kernel_size=3, strides=1, padding=padding, activation=activation)(conv)
+ print(conv.shape)
+
+ #self.logits = tf.keras.layers.Conv2D(1, kernel_size=1, strides=1, padding=padding, name='probability', activation=tf.nn.sigmoid)(conv)
+ self.logits = tf.keras.layers.Conv2D(1, kernel_size=1, strides=1, padding=padding, name='probability')(conv)
# conv = tf.nn.depth_to_space(conv, block_size=2)
# conv = tf.keras.layers.Activation(activation=activation)(conv)
@@ -755,8 +776,9 @@ class ESPCN:
pred = self.model([data])
self.test_probs = pred
+ pred = pred.numpy()
- return denormalize(pred, param, mean_std_dct[param])
+ return denormalize(pred, param, mean_std_dct)
def run(self, directory):
train_data_files = glob.glob(directory+'data_train*.npy')
@@ -780,10 +802,11 @@ class ESPCN:
def run_evaluate(self, nda_lr, param, ckpt_dir):
# self.setup_eval_pipeline(filename)
+ self.num_data_samples = 80000
self.build_model()
self.build_training()
self.build_evaluation()
- self.do_evaluate(nda_lr, param, ckpt_dir)
+ return self.do_evaluate(nda_lr, param, ckpt_dir)
if __name__ == "__main__":
--
GitLab