Update app.py

app.py

@@ -7,7 +7,6 @@ from PIL import Image as PILImage
 import io
 import os
 import base64
-import random
 
 def create_monitor_interface():
     api_key = os.getenv("GROQ_API_KEY")
@@ -18,9 +17,6 @@ def create_monitor_interface():
             self.model_name = "llama-3.2-90b-vision-preview"
             self.max_image_size = (800, 800)
             self.colors = [(0, 0, 255), (255, 0, 0), (0, 255, 0), (255, 255, 0), (255, 0, 255)]
-            self.last_analysis_time = 0
-            self.analysis_interval = 2
-            self.last_observations = []
             
         def resize_image(self, image):
             height, width = image.shape[:2]
@@ -38,7 +34,7 @@ def create_monitor_interface():
 
         def analyze_frame(self, frame: np.ndarray) -> str:
             if frame is None:
-                return "No frame received"
+                return ""
                 
             # Convert image
             if len(frame.shape) == 2:
@@ -66,10 +62,14 @@ def create_monitor_interface():
                             "content": [
                                 {
                                     "type": "text",
-                                    "text": """Analyze this image for safety hazards. For each issue, describe:
-                                    1. The location (top-left, center, bottom-right, etc.)
-                                    2. The specific safety concern
-                                    Format: - <location>position:description</location>"""
+                                    "text": """Analyze this image for safety concerns. For each specific issue you identify, provide:
+                                    1. Exact location in the image (e.g., 'top-left', 'center', 'bottom-right', etc.)
+                                    2. Description of the safety concern
+                                    
+                                    Format your response with each issue on a new line as:
+                                    - <location>position:detailed description of the safety concern</location>
+                                    
+                                    Be specific about what you observe in the image."""
                                 },
                                 {
                                     "type": "image_url",
@@ -84,16 +84,16 @@ def create_monitor_interface():
                             "content": ""
                         }
                     ],
-                    temperature=0.1,
-                    max_tokens=200,
+                    temperature=0.2,
+                    max_tokens=500,
                     top_p=1,
                     stream=False,
                     stop=None
                 )
                 return completion.choices[0].message.content
             except Exception as e:
-                print(f"Detailed error: {str(e)}")
-                return f"Analysis Error: {str(e)}"
+                print(f"Analysis error: {str(e)}")
+                return ""
 
         def get_region_coordinates(self, position: str, image_shape: tuple) -> tuple:
             height, width = image_shape[:2]
@@ -129,21 +129,24 @@ def create_monitor_interface():
                     position = parts[0]
                     description = ':'.join(parts[1:])
                 else:
-                    position = 'center'
-                    description = obs
+                    continue
                 
                 x1, y1, x2, y2 = self.get_region_coordinates(position, image.shape)
                 
+                # Draw rectangle
                 cv2.rectangle(image, (x1, y1), (x2, y2), color, 2)
                 
+                # Add label with background
                 label = description[:50] + "..." if len(description) > 50 else description
                 label_size = cv2.getTextSize(label, font, font_scale, thickness)[0]
                 
                 label_x = max(0, min(x1, width - label_size[0]))
                 label_y = max(20, y1 - 5)
                 
+                # Draw background for text
                 cv2.rectangle(image, (label_x, label_y - 20), 
                             (label_x + label_size[0], label_y), color, -1)
+                # Draw text
                 cv2.putText(image, label, (label_x, label_y - 5), 
                           font, font_scale, (255, 255, 255), thickness)
             
@@ -153,29 +156,26 @@ def create_monitor_interface():
             if frame is None:
                 return None, "No image provided"
             
-            current_time = time.time()
+            analysis = self.analyze_frame(frame)
             
-            if current_time - self.last_analysis_time >= self.analysis_interval:
-                analysis = self.analyze_frame(frame)
-                self.last_analysis_time = current_time
-                
-                observations = []
-                for line in analysis.split('\n'):
-                    line = line.strip()
-                    if line.startswith('-'):
-                        if '<location>' in line and '</location>' in line:
-                            start = line.find('<location>') + len('<location>')
-                            end = line.find('</location>')
-                            observation = line[start:end].strip()
-                            if observation:
-                                observations.append(observation)
-                
-                self.last_observations = observations
+            # Parse observations
+            observations = []
+            for line in analysis.split('\n'):
+                line = line.strip()
+                if line.startswith('-'):
+                    if '<location>' in line and '</location>' in line:
+                        start = line.find('<location>') + len('<location>')
+                        end = line.find('</location>')
+                        observation = line[start:end].strip()
+                        if observation and ':' in observation:
+                            observations.append(observation)
             
             display_frame = frame.copy()
-            annotated_frame = self.draw_observations(display_frame, self.last_observations)
-            
-            return annotated_frame, '\n'.join([f"- {obs}" for obs in self.last_observations])
+            if observations:
+                annotated_frame = self.draw_observations(display_frame, observations)
+                return annotated_frame, analysis
+            else:
+                return display_frame, "No safety concerns detected in the image."
 
     # Create the main interface
     monitor = SafetyMonitor()
@@ -185,9 +185,9 @@ def create_monitor_interface():
         
         with gr.Row():
             input_image = gr.Image(label="Upload Image")
-            output_image = gr.Image(label="Analysis")
+            output_image = gr.Image(label="Analysis Results")
         
-        analysis_text = gr.Textbox(label="Safety Concerns", lines=5)
+        analysis_text = gr.Textbox(label="Safety Analysis", lines=5)
             
         def analyze_image(image):
             if image is None:
@@ -207,9 +207,9 @@ def create_monitor_interface():
 
         gr.Markdown("""
         ## Instructions:
-        1. Upload an image to analyze safety concerns
-        2. View annotated results and detailed analysis
-        3. Each box highlights a potential safety issue
+        1. Upload an image to analyze
+        2. View identified safety concerns with bounding boxes
+        3. Read detailed analysis results
         """)
 
     return demo