revert to vision-agent code

app.py

@@ -4,11 +4,8 @@ import gradio as gr
 from typing import *
 from pillow_heif import register_heif_opener
 register_heif_opener()
-from PIL import Image
-import numpy as np
 import vision_agent as va
-from vision_agent.tools import owl_v2, overlay_bounding_boxes, save_image
-
+from vision_agent.tools import register_tool
 from vision_agent.tools import load_image, owl_v2, overlay_bounding_boxes, save_image
 
 from huggingface_hub import login
@@ -23,7 +20,7 @@ def detect_brain_tumor(image, debug: bool = False) -> str:
     Detects a brain tumor in the given image and saves the image with bounding boxes.
 
     Parameters:
-        image: The input image (can be PIL Image, numpy array, or file path).
+        image: The input image (as provided by Gradio).
         debug (bool): Flag to enable logging for debugging purposes.
 
     Returns:
@@ -32,28 +29,12 @@ def detect_brain_tumor(image, debug: bool = False) -> str:
     # Generate a unique output filename
     output_path = f"./output/tumor_detection_{int(time.time())}.jpg"
     
-    # Ensure image is in the correct format
-    if isinstance(image, str):
-        # If image is a file path
-        image = Image.open(image)
-    elif isinstance(image, np.ndarray):
-        # If image is already a numpy array
-        image = Image.fromarray(image)
-    elif not isinstance(image, Image.Image):
-        raise ValueError("Unsupported image type. Please provide a PIL Image, numpy array, or file path.")
-
-    # Convert to RGB if it's not
-    image = image.convert('RGB')
-    
-    # Convert PIL Image to numpy array for owl_v2
-    image_array = np.array(image)
-
     if debug:
-        print(f"Image loaded and converted to numpy array of shape {image_array.shape}")
+        print(f"Image received")
 
     # Step 2: Detect brain tumor using owl_v2
     prompt = "detect brain tumor"
-    detections = owl_v2(prompt, image_array)
+    detections = owl_v2(prompt, image)
     if debug:
         print(f"Detections: {detections}")
 
@@ -69,11 +50,6 @@ def detect_brain_tumor(image, debug: bool = False) -> str:
     
     return output_path
 
-# Example usage (uncomment to run):
-# detect_brain_tumor("/content/drive/MyDrive/kaggle/datasets/brain-tumor-image-dataset-semantic-segmentation_old/train_categories/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg", "/content/drive/MyDrive/kaggle/datasets/brain-tumor-image-dataset-semantic-segmentation_old/output/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg", debug=True)
-
-#########
-
 INTRO_TEXT="# 🔬🧠 CellVision AI -- Intelligent Cell Imaging Analysis 🤖🧫"
 IMAGE_PROMPT="Are these cells healthy or cancerous?"
 
@@ -82,7 +58,7 @@ with gr.Blocks(css="style.css") as demo:
     with gr.Tab("Agentic Detection"):
         with gr.Row():
             with gr.Column():
-                image = gr.Image(type="pil")
+                image = gr.Image(type="numpy")
             with gr.Column():
                 text_input = gr.Text(label="Input Text")
                 text_output = gr.Text(label="Text Output")
@@ -100,21 +76,19 @@ with gr.Blocks(css="style.css") as demo:
             outputs=chat_outputs,
         )
         
-        examples = [["./examples/194_jpg.rf.3e3dd592d034bb5ee27a978553819f42.jpg", "./output/194_jpg.rf.3e3dd592d034bb5ee27a978553819f42.jpg"],
-                    ["./examples/239_jpg.rf.3dcc0799277fb78a2ab21db7761ccaeb.jpg", "./output/239_jpg.rf.3dcc0799277fb78a2ab21db7761ccaeb.jpg"],
-                    ["./examples/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg", "./output/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg"],
-                    ["./examples/1491_jpg.rf.3c658e83538de0fa5a3f4e13d7d85f12.jpg", "./output/1491_jpg.rf.3c658e83538de0fa5a3f4e13d7d85f12.jpg"],
-                    ["./examples/1550_jpg.rf.3d067be9580ec32dbee5a89c675d8459.jpg", "./output/1550_jpg.rf.3d067be9580ec32dbee5a89c675d8459.jpg"],
-                    ["./examples/2256_jpg.rf.3afd7903eaf3f3c5aa8da4bbb928bc19.jpg", "./output/2256_jpg.rf.3afd7903eaf3f3c5aa8da4bbb928bc19.jpg"],
-                    ["./examples/2871_jpg.rf.3b6eadfbb369abc2b3bcb52b406b74f2.jpg", "./output/2871_jpg.rf.3b6eadfbb369abc2b3bcb52b406b74f2.jpg"],
-                    ["./examples/2921_jpg.rf.3b952f91f27a6248091e7601c22323ad.jpg", "./output/2921_jpg.rf.3b952f91f27a6248091e7601c22323ad.jpg"],
+        examples = [["./examples/194_jpg.rf.3e3dd592d034bb5ee27a978553819f42.jpg"],
+                    ["./examples/239_jpg.rf.3dcc0799277fb78a2ab21db7761ccaeb.jpg"],
+                    ["./examples/1385_jpg.rf.3c67cb92e2922dba0e6dba86f69df40b.jpg"],
+                    ["./examples/1491_jpg.rf.3c658e83538de0fa5a3f4e13d7d85f12.jpg"],
+                    ["./examples/1550_jpg.rf.3d067be9580ec32dbee5a89c675d8459.jpg"],
+                    ["./examples/2256_jpg.rf.3afd7903eaf3f3c5aa8da4bbb928bc19.jpg"],
+                    ["./examples/2871_jpg.rf.3b6eadfbb369abc2b3bcb52b406b74f2.jpg"],
+                    ["./examples/2921_jpg.rf.3b952f91f27a6248091e7601c22323ad.jpg"],
                     ]
         gr.Examples(
             examples=examples,
             inputs=chat_inputs,
         )
 
-#########
-
 if __name__ == "__main__":
     demo.queue(max_size=10).launch(debug=True)