changed structure and UI

app.py

@@ -97,7 +97,7 @@ def predict(img, caption):
         caption = get_caption(img)
     print("New caption-->",caption)
     prediction=[]
-    models_list = ['clip-sd.pth', 'clip-glide.pth', 'clip-ld.pth']
+    models_list = ['models/clip-sd.pth', 'models/clip-glide.pth', 'models/clip-ld.pth']
     modality = "Image+Text"
     for model_path in models_list:
         model = get_model(model_path, modality)
@@ -108,18 +108,20 @@ def predict(img, caption):
             print('------------>', out)
             prediction.append(out.item())
 
-    if prediction[0] > 0.5:
-        return "Fake"
+    # Count the number of predictions that are greater than or equal to 0.5
+    count_ones = sum(1 for p in prediction if p >= 0.5)
+    if count_ones > len(prediction) / 2:
+        return "Fake Image"
     else:
-        return "Real"
+        return "Real Image  "
 
 
 # Create Gradio interface
-image_input = gr.Image(type="pil", label="Upload Image")
+image_input = gr.Image(type="pil", label="Input Image")
 text_input = gr.Textbox(label="Caption for image (Optional)")
 
 iface = gr.Interface(fn=predict,
              inputs=[image_input, text_input],
              outputs=gr.Label(),
-             examples=[["trump-fake.jpeg", "Donald Trump being arrested by authorities."]])
+             examples=[["examples/trump-fake.jpeg", "Donald Trump being arrested by authorities."]])
 iface.launch()

models/clip-glide.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3492e0892487a66dfda34e36aa407a2765b6d50bdf72ab50b6543d1541c4982e
+size 425232242

models/clip-ld.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e66b78073bb830105d52dd0040eba2706287c7637a1959363a801c5b2f9d1cc9
+size 425230745

models/clip-sd.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b062950f01721fd4e947cc0ce1f0a7ae45a45dfbc94d6a0688ca2d0c650d83a
+size 425230745