added options

app.py

@@ -34,12 +34,16 @@ model.load_state_dict(state['val_model_dict'])
 config = resolve_data_config({}, model=vit)
 config['no_aug'] = True
 config['interpolation'] = 'bilinear'
-config['crop_pct'] = 1
-config['crop_mode'] = 'border'
-transform = create_transform(**config)
 
 # Inference function
-def query_image(input_img, query, binarize, eval_threshold):
+def query_image(input_img, query, binarize, eval_threshold, crop_mode, crop_pct):
+
+    if crop_mode == 'center':
+        crop_mode = None
+
+    config['crop_pct'] = crop_pct
+    config['crop_mode'] = crop_mode
+    transform = create_transform(**config)
 
     PIL_image = Image.fromarray(input_img, "RGB")
     img = transform(PIL_image)
@@ -92,7 +96,8 @@ the model will binarize each propability based on set eval_threshold.
 demo = gr.Interface(
     query_image, 
     #inputs=[gr.Image(), "text", "checkbox", gr.Slider(0, 1, value=0.25)],
-    inputs=[gr.Image(type='numpy', label='input_img').style(height=340, width=600), "text", "checkbox", gr.Slider(0, 1, value=0.25)],  
+    inputs=[gr.Image(type='numpy', label='input_img').style(height=270, width=600), "text", "checkbox", gr.Slider(0, 1, value=0.25), 
+            gr.inputs.Dropdown("center", "squash", "border", label='crop_mode'), gr.Slider(0, 1, value=1)],  
     #outputs="image",
     outputs=gr.Image(type='numpy', label='output').style(height=600, width=600),
     title="Object Detection Using Textual Queries",