fix apps

ppgan/apps/base_predictor.py

@@ -62,13 +62,13 @@ class BasePredictor(object):
 
         return out
 
-    def is_video(self, input):
+    def is_image(self, input):
         try:
             img = Image.open(input)
             _ = img.size
-            return False
-        except:
             return True
+        except:
+            return False
 
     def run(self):
         raise NotImplementedError

ppgan/apps/deoldify_predictor.py

@@ -128,7 +128,7 @@ class DeOldifyPredictor(BasePredictor):
         return frame_pattern_combined, vid_out_path
 
     def run(self, input):
-        if self.is_video(input):
+        if not self.is_image(input):
             return self.run_video(input)
         else:
             pred_img = self.run_image(input)

ppgan/apps/realsr_predictor.py

@@ -101,7 +101,7 @@ class RealSRPredictor(BasePredictor):
         if not os.path.exists(self.output):
             os.makedirs(self.output)
 
-        if self.is_video(input):
+        if not self.is_image(input):
             return self.run_video(input)
         else:
             pred_img = self.run_image(input)

ppgan/models/generators/deoldify.py

@@ -2,10 +2,9 @@ import numpy as np
 import paddle
 import paddle.nn as nn
 import paddle.nn.functional as F
+from paddle.vision.models import resnet101
 
 from .hook import hook_outputs, model_sizes, dummy_eval
-from ..backbones import resnet34, resnet101
-# from paddle.vision.models import resnet101
 from ...modules.nn import Spectralnorm
 
 
@@ -273,7 +272,7 @@ class PixelShuffle_ICNR(nn.Layer):
 
         self.shuf = PixelShuffle(scale)
 
-        self.pad = ReplicationPad2d((1, 0, 1, 0))
+        self.pad = ReplicationPad2d([1, 0, 1, 0])
         self.blur = nn.Pool2D(2, pool_stride=1, pool_type='avg')
         self.relu = relu(True, leaky=leaky)
 
@@ -339,7 +338,7 @@ class CustomPixelShuffle_ICNR(nn.Layer):
 
         self.shuf = PixelShuffle(scale)
 
-        self.pad = ReplicationPad2d((1, 0, 1, 0))
+        self.pad = ReplicationPad2d([1, 0, 1, 0])
         self.blur = nn.Pool2D(2, pool_stride=1, pool_type='avg')
         self.relu = nn.LeakyReLU(
             leaky) if leaky is not None else nn.ReLU()  #relu(True, leaky=leaky)
@@ -410,7 +409,7 @@ class ReplicationPad2d(nn.Layer):
         self.size = size
 
     def forward(self, x):
-        return F.pad2d(x, self.size, mode="edge")
+        return F.pad(x, self.size, mode="replicate")
 
 
 def conv1d(ni: int,
@@ -420,7 +419,7 @@ def conv1d(ni: int,
            padding: int = 0,
            bias: bool = False):
     "Create and initialize a `nn.Conv1d` layer with spectral normalization."
-    conv = nn.Conv1d(ni, no, ks, stride=stride, padding=padding, bias_attr=bias)
+    conv = nn.Conv1D(ni, no, ks, stride=stride, padding=padding, bias_attr=bias)
     return Spectralnorm(conv)
 
 

ppgan/models/generators/hook.py

@@ -77,7 +77,7 @@ class Hooks():
 
 def _hook_inner(m, i, o):
     return o if isinstance(
-        o, paddle.framework.Variable) else o if is_listy(o) else list(o)
+        o, paddle.fluid.framework.Variable) else o if is_listy(o) else list(o)
 
 
 def hook_output(module, detach=True, grad=False):

ppgan/models/generators/mobile_resnet.py

@@ -109,9 +109,9 @@ class MobileResnetBlock(nn.Layer):
 
         p = 0
         if self.padding_type == 'reflect':
-            self.conv_block.extend([nn.ReflectionPad2d([1, 1, 1, 1])])
+            self.conv_block.extend([nn.Pad2D([1, 1, 1, 1], mode='reflect')])
         elif self.padding_type == 'replicate':
-            self.conv_block.extend([nn.ReplicationPad2d([1, 1, 1, 1])])
+            self.conv_block.extend([nn.Pad2D([1, 1, 1, 1], mode='replicate')])
         elif self.padding_type == 'zero':
             p = 1
         else:

ppgan/models/generators/remaster.py

@@ -128,7 +128,7 @@ class NetworkR(nn.Layer):
         super(NetworkR, self).__init__()
 
         self.layers = nn.Sequential(
-            nn.ReplicationPad3d((1, 1, 1, 1, 1, 1)),
+            nn.Pad3D((1, 1, 1, 1, 1, 1), mode='replicate'),
             TempConv(1,
                      64,
                      kernel_size=(3, 3, 3),
@@ -165,7 +165,7 @@ class NetworkC(nn.Layer):
         super(NetworkC, self).__init__()
 
         self.down1 = nn.Sequential(
-            nn.ReplicationPad3d((1, 1, 1, 1, 0, 0)),
+            nn.Pad3D((1, 1, 1, 1, 0, 0), mode='replicate'),
             TempConv(1, 64, stride=(1, 2, 2), padding=(0, 0, 0)),
             TempConv(64, 128), TempConv(128, 128),
             TempConv(128, 256, stride=(1, 2, 2)), TempConv(256, 256),