update

README.md

@@ -1,7 +1,7 @@
 # DeepMindBreak
 Decensoring Hentai with Deep Neural Networks
 
-
+This project applies an implementation of [Globally and Locally Consistent Image Completion](http://hi.cs.waseda.ac.jp/%7Eiizuka/projects/completion/data/completion_sig2017.pdf) to the problem of hentai decensorship. Using a deep fully convolutional neural network, DeepMindBreak can replace censored artwork in hentai with plausible reconstructions. The user needs to only specify the censored regions for the algorithm to run.
 
 # Limitations
 
@@ -15,9 +15,10 @@ It does NOT work with:
 - Real life porn
 - Mosaic censorship
 - Censorship of nipples
-- Complete censorship
 - Animated gifs/videos
 
+In particular, if a vagina or penis is completely censored out, THERE IS NO HOPE OF RECOVERY.
+
 # Dependencies
 
 - Python 2/3

src/decensor.py

@@ -30,7 +30,7 @@ def test():
     sess.run(init_op)
 
     saver = tf.train.Saver()
-    saver.restore(sess, '../saved_models/latest')
+    saver.restore(sess, '/saved_models/latest')
 
     x_test = np.load(test_npy)
     np.random.shuffle(x_test)

src/test.py

@@ -30,7 +30,7 @@ def test():
     sess.run(init_op)
 
     saver = tf.train.Saver()
-    saver.restore(sess, '../saved_models/latest')
+    saver.restore(sess, '/saved_models/latest')
 
     x_test = np.load(test_npy)
     np.random.shuffle(x_test)
@@ -50,7 +50,7 @@ def test():
             masked = raw * (1 - mask_batch[i]) + np.ones_like(raw) * mask_batch[i] * 255
             img = completion[i]
             img = np.array((img + 1) * 127.5, dtype=np.uint8)
-            dst = './output/{}.jpg'.format("{0:06d}".format(cnt))
+            dst = '/output/{}.jpg'.format("{0:06d}".format(cnt))
             output_image([['Input', masked], ['Output', img], ['Ground Truth', raw]], dst)
 
 

src/train.py

@@ -33,9 +33,9 @@ def train():
     init_op = tf.global_variables_initializer()
     sess.run(init_op)
 
-    if tf.train.get_checkpoint_state('./saved_model'):
+    if tf.train.get_checkpoint_state('/saved_model'):
         saver = tf.train.Saver()
-        saver.restore(sess, './saved_model/latest')
+        saver.restore(sess, '/saved_model/latest')
 
     x_train, x_test = load.load()
     x_train = np.array([a / 127.5 - 1 for a in x_train])
@@ -69,9 +69,9 @@ def train():
 
 
             saver = tf.train.Saver()
-            saver.save(sess, './saved_model/latest', write_meta_graph=False)
+            saver.save(sess, '/saved_model/latest', write_meta_graph=False)
             if sess.run(epoch) == PRETRAIN_EPOCH:
-                saver.save(sess, './saved_model/pretrained', write_meta_graph=False)
+                saver.save(sess, '/saved_model/pretrained', write_meta_graph=False)
 
 
         # Discrimitation
@@ -106,10 +106,10 @@ def train():
             x_batch = x_test[:BATCH_SIZE]
             completion = sess.run(model.completion, feed_dict={x: x_batch, mask: mask_batch, is_training: False})
             sample = np.array((completion[0] + 1) * 127.5, dtype=np.uint8)
-            cv2.imwrite('./output/{}.jpg'.format("{0:06d}".format(sess.run(epoch))), cv2.cvtColor(sample, cv2.COLOR_RGB2BGR))
+            cv2.imwrite('/output/{}.jpg'.format("{0:06d}".format(sess.run(epoch))), cv2.cvtColor(sample, cv2.COLOR_RGB2BGR))
 
             saver = tf.train.Saver()
-            saver.save(sess, './saved_model/latest', write_meta_graph=False)
+            saver.save(sess, '/saved_model/latest', write_meta_graph=False)
 
 
 def get_points():