fix: updated typo

AppGuiFunctions.py

@@ -1,3 +1,5 @@
+import os
+
 import cv2
 import numpy as np
 import shutil
@@ -27,37 +29,44 @@ def saveFile(self, savepath):
 
     def imageFitter(self, file_input_path, file_output_path, winwidth, winheight):
         rawimg = cv2.imread(file_input_path, 1)
+        if rawimg is None:
+            raise FileNotFoundError(f"Could not read image: {file_input_path}")
 
         self.imgrealwidth = rawimg.shape[1]
         self.imgrealheight = rawimg.shape[0]
-        ##Will change the logic below later, for now its been configured for square window shapes :P
+        # Will change the logic below later, for now it's been configured for square window shapes
 
-        if rawimg.shape[1] >= rawimg.shape[0] :
+        if rawimg.shape[1] >= rawimg.shape[0]:
             newx = winwidth
-            newy = (rawimg.shape[0]/rawimg.shape[1])*winwidth
+            newy = (rawimg.shape[0] / rawimg.shape[1]) * winwidth
         else:
             newy = winheight
-            newx = (rawimg.shape[1]/rawimg.shape[0])*winheight
-        
+            newx = (rawimg.shape[1] / rawimg.shape[0]) * winheight
+
         newdim = (int(newx), int(newy))
 
-        resizedimg = cv2.resize(rawimg, newdim, interpolation = cv2.INTER_AREA)
+        resizedimg = cv2.resize(rawimg, newdim, interpolation=cv2.INTER_AREA)
+        # Ensure output directory exists
+        os.makedirs(os.path.dirname(os.path.abspath(file_output_path)), exist_ok=True)
         cv2.imwrite(file_output_path, resizedimg)
 
     def superresUpscaler(self, model):
         self.finaloutputpath = "processimage/finalprocessedimage_" + model + ".jpg"
         self.displayoutput = "processimage/downscaled_finalprocessedimage_" + model + ".jpg"
-        
+
         self.ganPredict = DeepLearningGANModels(model, self.path_to_file)
         model_img = self.ganPredict.modelBash()
 
-        array_img = np.array(model_img).astype(np.float32)
+        if model_img is None:
+            return False
+
+        # Ensure output directory exists
+        os.makedirs(os.path.dirname(os.path.abspath(self.finaloutputpath)), exist_ok=True)
+
+        array_img = np.array(model_img).astype(np.uint8)
         processed_model_img = cv2.cvtColor(array_img, cv2.COLOR_RGB2BGR)
         cv2.imwrite(self.finaloutputpath, processed_model_img)
 
         self.imageFitter(self.finaloutputpath, self.displayoutput, self.winwidth, self.winheight)
 
-        if model_img is not None :
-            return True
-        else :
-            return False
+        return True

AppWindowQt.py

@@ -1,8 +1,10 @@
 # -*- coding: utf-8 -*-
+import os
+import sys
+
 from PyQt5 import QtCore, QtGui, QtWidgets
 from PyQt5.QtWidgets import QFileDialog, QMessageBox, QLabel, QMainWindow
 from AppGuiFunctions import GuiFunctions
-import sys
 
 
 class QLabel_alterada(QLabel):
@@ -167,14 +169,21 @@ def retranslateUi(self, MainWindow):
         self.actionSave.setShortcut(_translate("MainWindow", "Ctrl+S"))
 
     def processImage(self, model):
-        if not self.file_has_been_loaded :
+        if not self.file_has_been_loaded:
             self.showNoFilePopup()
-        else :
-            successRes = self.gui_functions.superresUpscaler(model)
-            if successRes :
-                self.processed_photo.setPixmap(QtGui.QPixmap("processimage/downscaled_finalprocessedimage_" + model + ".jpg"))
+        else:
+            try:
+                successRes = self.gui_functions.superresUpscaler(model)
+            except Exception:
+                self.showFatalPopup()
+                return
+
+            if successRes:
+                self.processed_photo.setPixmap(QtGui.QPixmap(
+                    "processimage/downscaled_finalprocessedimage_" + model + ".jpg"
+                ))
                 self.photo_has_been_processed = True
-            else :
+            else:
                 self.showFatalPopup()
 
     def displayChosenImage(self):
@@ -189,32 +198,38 @@ def openExtraWindow(self):
 
     def openFileDialogBox(self):
         file_path = QFileDialog.getOpenFileName()[0]
-        extension = file_path.split('.')[-1] ##Use this for checking filetype and throwing errors and stuff
-
-        if extension in self.accepted_extensions :
+        if not file_path:
+            return  # user cancelled
+
+        _, ext = os.path.splitext(file_path)
+        extension = ext.lstrip(".").lower()
+
+        if extension in self.accepted_extensions:
             self.gui_functions.openFile(file_path, self.imgwidth, self.imgheight)
             self.file_has_been_loaded = True
             self.displayChosenImage()
 
             self.processed_photo.setText(QtCore.QCoreApplication.translate("MainWindow", "Click Here to View Your UpScaled Image"))
             self.photo_has_been_processed = False
-        elif extension == "" :
-            pass
-        else :
+        else:
             self.showExtensionsMismatchPopup(extension)
 
     def saveFileDialogBox(self):
-        if not self.file_has_been_loaded :
+        if not self.file_has_been_loaded:
             self.showNoFilePopup()
-        else :
-            file_path = QFileDialog.getSaveFileName()[0]
-            extension = file_path.split('.')[-1] ##Use this for checking filetype and throwing errors and stuff
-
-            if extension == file_path :
-                file_path = file_path + ".jpg"
-                self.gui_functions.saveFile(file_path)
-            else :
-                self.gui_functions.saveFile(file_path)
+            return
+        if not self.photo_has_been_processed:
+            self.showNoUpscalingDonePopup()
+            return
+
+        file_path = QFileDialog.getSaveFileName()[0]
+        if not file_path:
+            return  # user cancelled
+
+        if not os.path.splitext(file_path)[1]:
+            file_path = file_path + ".jpg"
+
+        self.gui_functions.saveFile(file_path)
 
     def showNoFilePopup(self):
         popmsg = QMessageBox()

EnhancerGANModel.py

@@ -16,7 +16,8 @@ def __init__(self, model, input_path):
         self.input_path = input_path
 
     def load_image(self, path):
-        return np.array(Image.open(path))[:,:,:3]
+        img = Image.open(path).convert("RGB")
+        return np.array(img)
 
     def modelBash(self):
         if self.modelname == "srganx4" :
@@ -27,23 +28,31 @@ def modelBash(self):
             return None
 
     def wdsr_b_finetuned(self):
-        weights_dir = 'model/weights/wdsr'
+        weights_dir = os.path.join(os.path.dirname(__file__), 'model', 'weights', 'wdsr')
+        weights_path = os.path.join(weights_dir, 'weights-wdsr-b-fine-tuned.h5')
+        if not os.path.isfile(weights_path):
+            raise FileNotFoundError(
+                f"WDSR weight file not found: {weights_path}\n"
+                "Download weights from https://drive.google.com/drive/folders/..."
+            )
+
         wdsr_fine_tuned = wdsr_b(scale=4, num_res_blocks=32)
-        wdsr_fine_tuned.load_weights(os.path.join(weights_dir, 'weights-wdsr-b-fine-tuned.h5'))
+        wdsr_fine_tuned.load_weights(weights_path)
 
         lr = self.load_image(self.input_path)
-        gan_wdsr_finetuned = resolve_single(wdsr_fine_tuned, lr)
-
-        return gan_wdsr_finetuned
+        return resolve_single(wdsr_fine_tuned, lr)
 
     def srganx4Process(self):
-        weights_dir = 'model/weights/srgan'
-        weights_file = lambda filename: os.path.join(weights_dir, filename)
+        weights_dir = os.path.join(os.path.dirname(__file__), 'model', 'weights', 'srgan')
+        weights_path = os.path.join(weights_dir, 'gan_generator.h5')
+        if not os.path.isfile(weights_path):
+            raise FileNotFoundError(
+                f"SRGAN weight file not found: {weights_path}\n"
+                "Download weights from https://drive.google.com/drive/folders/..."
+            )
 
         gan_generator = srgangenerator()
-        gan_generator.load_weights(weights_file('gan_generator.h5'))
+        gan_generator.load_weights(weights_path)
 
         lr = self.load_image(self.input_path)
-        gan_sr = resolve_single(gan_generator, lr)
-
-        return gan_sr
+        return resolve_single(gan_generator, lr)

model/WDSRFineTuned.py

@@ -1,10 +1,13 @@
-import tensorflow_addons as tfa
-
-from tensorflow.python.keras.layers import Add, Conv2D, Input, Lambda
-from tensorflow.python.keras.models import Model
+from tensorflow.keras.layers import Add, Conv2D, Input, Lambda
+from tensorflow.keras.models import Model
 
 from model.common import normalize, denormalize, pixel_shuffle
 
+try:
+    import tensorflow_addons as tfa
+except ImportError:
+    tfa = None
+
 
 def wdsr_a(scale, num_filters=32, num_res_blocks=8, res_block_expansion=4, res_block_scaling=None):
     return wdsr(scale, num_filters, num_res_blocks, res_block_expansion, res_block_scaling, res_block_a)
@@ -56,4 +59,12 @@ def res_block_b(x_in, num_filters, expansion, kernel_size, scaling):
 
 
 def conv2d_weightnorm(filters, kernel_size, padding='same', activation=None, **kwargs):
-    return tfa.layers.WeightNormalization(Conv2D(filters, kernel_size, padding=padding, activation=activation, **kwargs), data_init=False)
+    if tfa is None:
+        raise ImportError(
+            "tensorflow-addons is required for WDSR models. "
+            "Install with: pip install tensorflow-addons"
+        )
+    return tfa.layers.WeightNormalization(
+        Conv2D(filters, kernel_size, padding=padding, activation=activation, **kwargs),
+        data_init=False,
+    )

model/srgan.py

@@ -1,6 +1,6 @@
-from tensorflow.python.keras.layers import Add, BatchNormalization, Conv2D, Dense, Flatten, Input, LeakyReLU, PReLU, Lambda
-from tensorflow.python.keras.models import Model
-from tensorflow.python.keras.applications.vgg19 import VGG19
+from tensorflow.keras.layers import Add, BatchNormalization, Conv2D, Dense, Flatten, Input, LeakyReLU, PReLU, Lambda
+from tensorflow.keras.models import Model
+from tensorflow.keras.applications.vgg19 import VGG19
 
 from model.common import pixel_shuffle, normalize_01, normalize_m11, denormalize_m11
 

run.py

@@ -1,9 +1,25 @@
-from AppWindowQt import Ui_MainWindow
-import tensorflow as tf
 import os
+import sys
+
+from AppWindowQt import Ui_MainWindow
+
+
+def main():
+    # Configure GPU memory growth if a GPU is available.
+    try:
+        import tensorflow as tf
+
+        physical_devices = tf.config.list_physical_devices("GPU")
+        if physical_devices:
+            tf.config.experimental.set_memory_growth(physical_devices[0], True)
+    except ImportError:
+        pass  # TensorFlow not installed — CPU fallback or inference-only
+    except Exception:
+        pass  # GPU config failed — continue anyway
+
+    main_ui = Ui_MainWindow()
+    main_ui.execute()
 
-physical_devices = tf.config.list_physical_devices('GPU')
-tf.config.experimental.set_memory_growth(physical_devices[0], True)
 
-main_ui = Ui_MainWindow()
-main_ui.execute()
+if __name__ == "__main__":
+    main()