Update app.py

app.py

@@ -18,25 +18,29 @@ os.system("wget https://huggingface.co/spaces/codeslake/RefVSR/resolve/main/RefV
 
 sys.path.append("RefVSR")
 
-## Input setup (creates folders and places inputs corresponding to the original RefVSR code)
+## I/O setup (creates folders and places inputs corresponding to the original RefVSR code)
+# HD input
 HR_LR_path = "test/RealMCVSR/test/HR/UW/0000"
 HR_Ref_path = "test/RealMCVSR/test/HR/W/0000"
 HR_Ref_path_T = "test/RealMCVSR/test/HR/T/0000"
+os.makedirs(HR_LR_path)
+os.makedirs(HR_Ref_path)
+os.makedirs(HR_Ref_path_T)
+os.system("wget https://www.dropbox.com/s/xv6inxwy0so4ni0/LR.png -O HR_LR.png")
+os.system("wget https://www.dropbox.com/s/abydd1oczs1163l/Ref.png -O HR_Ref.png")
+
+# 4x downsampled input
 LR_path = "test/RealMCVSR/test/LRx4/UW/0000"
 Ref_path = "test/RealMCVSR/test/LRx4/W/0000"
 Ref_path_T = "test/RealMCVSR/test/LRx4/T/0000"
 os.makedirs(LR_path)
 os.makedirs(Ref_path)
 os.makedirs(Ref_path_T)
-os.makedirs(HR_LR_path)
-os.makedirs(HR_Ref_path)
-os.makedirs(HR_Ref_path_T)
-os.makedirs('result')
 os.system("wget https://www.dropbox.com/s/vqekqdz80d85gi4/UW.png -O LR.png")
 os.system("wget https://www.dropbox.com/s/lsopmquhpm87v83/W.png -O Ref.png")
-os.system("wget https://www.dropbox.com/s/xv6inxwy0so4ni0/LR.png -O HR_LR.png")
-os.system("wget https://www.dropbox.com/s/abydd1oczs1163l/Ref.png -O HR_Ref.png")
 
+# output directory
+os.makedirs('result')
 
 ## resize if necessary (not used)
 def resize(img):
@@ -50,9 +54,11 @@ def resize(img):
         img = img.resize((wsize,hsize), Image.ANTIALIAS)
     return img
 
-#################### 8K ##################
+
+    
+#################### low res ##################
 ## inference  
-def inference_8K(LR, Ref):
+def inference(LR, Ref):
     ## resize for user selected input (not used)
     #LR = resize(LR)
     #Ref = resize(Ref)
@@ -67,35 +73,35 @@ def inference_8K(LR, Ref):
     
     ## Run RefVSR model
     os.system("python -B run.py \
-                --mode RefVSR_MFID_8K \
-                --config config_RefVSR_MFID_8K \
+                --mode RefVSR_MFID \
+                --config config_RefVSR_MFID \
                 --data RealMCVSR \
-                --ckpt_abs_name ckpt/RefVSR_MFID_8K.pytorch \
+                --ckpt_abs_name ckpt/RefVSR_MFID.pytorch \
                 --data_offset ./test \
                 --output_offset ./result \
                 --qualitative_only \
                 --cpu \
-                --is_gradio")               
+                --is_gradio")             
     return "result/0000.png"
-    
+
 title="RefVSR"
-description="Demo application for Reference-based Video Super-Resolution (RefVSR). Upload a low-resolution frame and a reference frame to 'LR' and 'Ref' input windows, respectively. The demo runs on CPUs and takes about 120s."
+description="Demo application for Reference-based Video Super-Resolution (RefVSR). Upload a low-resolution frame and a reference frame to 'LR' and 'Ref' input windows, respectively. The demo runs on CPUs and takes about 150s."
 
-article = "<p style='text-align: center'><b>To check the full capability of the module, we recommend to clone Github repository and run RefVSR models on videos using GPUs.</b></p><p style='text-align: center'>This demo runs on CPUs and only supports RefVSR for a single LR and Ref frame due to computational complexity. Hence, the model will not take advantage of temporal LR and Ref frames.</p><p style='text-align: center'>The model is trained by the proposed two-stage training strategy. The sample frames are in HD resolution (1920x1080) and saved in the PNG format. </p><p style='text-align: center'><a href='https://junyonglee.me/projects/RefVSR' target='_blank'>Project</a> | <a href='https://arxiv.org/abs/2203.14537' target='_blank'>arXiv</a> | <a href='https://github.com/codeslake/RefVSR' target='_blank'>Github</a></p>"
+article = "<p style='text-align: center'><b>To check the full capability of the module, we recommend to clone Github repository and run RefVSR models on videos using GPUs.</b></p><p style='text-align: center'>This demo runs on CPUs and only supports RefVSR for a single LR and Ref frame due to computational complexity. Hence, the model will not take advantage of temporal LR and Ref frames.</p><p style='text-align: center'>The model is trained by the proposed pre-training strategy only. The sample frames are in 430x270 resolution and saved in the PNG format. </p><p style='text-align: center'><a href='https://junyonglee.me/projects/RefVSR' target='_blank'>Project</a> | <a href='https://arxiv.org/abs/2203.14537' target='_blank'>arXiv</a> | <a href='https://github.com/codeslake/RefVSR' target='_blank'>Github</a></p>"
 
 ## resize for sample (not used)
 #LR = resize(Image.open('LR.png')).save('LR.png')
 #Ref = resize(Image.open('Ref.png')).save('Ref.png')
 
 ## input
-examples=[['HR_LR.png', 'HR_Ref.png']]
+examples=[['LR.png', 'Ref.png']]
 
 ## interface
-gr.Interface(inference_8K,[gr.inputs.Image(type="pil"), gr.inputs.Image(type="pil")],gr.outputs.Image(type="file"),title=title,description=description,article=article,theme ="peach",examples=examples).launch(enable_queue=True)    
-    
-#################### low res ##################
+gr.Interface(inference,[gr.inputs.Image(type="pil"), gr.inputs.Image(type="pil")],gr.outputs.Image(type="file"),title=title,description=description,article=article,theme ="peach",examples=examples).launch(enable_queue=True)
+
+#################### 8K ##################
 ## inference  
-def inference(LR, Ref):
+def inference_8K(LR, Ref):
     ## resize for user selected input (not used)
     #LR = resize(LR)
     #Ref = resize(Ref)
@@ -110,29 +116,28 @@ def inference(LR, Ref):
     
     ## Run RefVSR model
     os.system("python -B run.py \
-                --mode RefVSR_MFID \
-                --config config_RefVSR_MFID \
+                --mode RefVSR_MFID_8K \
+                --config config_RefVSR_MFID_8K \
                 --data RealMCVSR \
-                --ckpt_abs_name ckpt/RefVSR_MFID.pytorch \
+                --ckpt_abs_name ckpt/RefVSR_MFID_8K.pytorch \
                 --data_offset ./test \
                 --output_offset ./result \
                 --qualitative_only \
                 --cpu \
-                --is_gradio")             
+                --is_gradio")               
     return "result/0000.png"
-
+    
 title="RefVSR"
-description="Demo application for Reference-based Video Super-Resolution (RefVSR). Upload a low-resolution frame and a reference frame to 'LR' and 'Ref' input windows, respectively. The demo runs on CPUs and takes about 150s."
+description="Demo application for Reference-based Video Super-Resolution (RefVSR). Upload a low-resolution frame and a reference frame to 'LR' and 'Ref' input windows, respectively. The demo runs on CPUs and takes about 120s."
 
-article = "<p style='text-align: center'><b>To check the full capability of the module, we recommend to clone Github repository and run RefVSR models on videos using GPUs.</b></p><p style='text-align: center'>This demo runs on CPUs and only supports RefVSR for a single LR and Ref frame due to computational complexity. Hence, the model will not take advantage of temporal LR and Ref frames.</p><p style='text-align: center'>The model is trained by the proposed pre-training strategy only. The sample frames are in 430x270 resolution and saved in the PNG format. </p><p style='text-align: center'><a href='https://junyonglee.me/projects/RefVSR' target='_blank'>Project</a> | <a href='https://arxiv.org/abs/2203.14537' target='_blank'>arXiv</a> | <a href='https://github.com/codeslake/RefVSR' target='_blank'>Github</a></p>"
+article = "<p style='text-align: center'><b>To check the full capability of the module, we recommend to clone Github repository and run RefVSR models on videos using GPUs.</b></p><p style='text-align: center'>This demo runs on CPUs and only supports RefVSR for a single LR and Ref frame due to computational complexity. Hence, the model will not take advantage of temporal LR and Ref frames.</p><p style='text-align: center'>The model is trained by the proposed two-stage training strategy. The sample frames are in HD resolution (1920x1080) and saved in the PNG format. </p><p style='text-align: center'><a href='https://junyonglee.me/projects/RefVSR' target='_blank'>Project</a> | <a href='https://arxiv.org/abs/2203.14537' target='_blank'>arXiv</a> | <a href='https://github.com/codeslake/RefVSR' target='_blank'>Github</a></p>"
 
 ## resize for sample (not used)
 #LR = resize(Image.open('LR.png')).save('LR.png')
 #Ref = resize(Image.open('Ref.png')).save('Ref.png')
 
 ## input
-examples=[['LR.png', 'Ref.png']]
+examples=[['HR_LR.png', 'HR_Ref.png']]
 
 ## interface
-gr.Interface(inference,[gr.inputs.Image(type="pil"), gr.inputs.Image(type="pil")],gr.outputs.Image(type="file"),title=title,description=description,article=article,theme ="peach",examples=examples).launch(enable_queue=True)
-
+gr.Interface(inference_8K,[gr.inputs.Image(type="pil"), gr.inputs.Image(type="pil")],gr.outputs.Image(type="file"),title=title,description=description,article=article,theme ="peach",examples=examples).launch(enable_queue=True)