Update app.py

app.py

@@ -70,106 +70,101 @@ def extract_frames_from_video(video, frame_points=[0, 0.5, 1], max_size=(800, 80
     cap.release()
     return frames
 
-def analyze_mixed_input(input_files):
-    if not input_files:
-        logger.warning("No input files provided")
-        return [("No input", "Error: Please upload images or a video for analysis.")]
-
+def analyze_file(file):
+    """Analyze a single file (image or video)"""
     try:
-        logger.info("Starting analysis")
-        results = []
-
-        instruction = ("You are an AI assistant specialized in analyzing images for safety issues. "
-                       "Your task is first to explain what you see in the image and determine if the image shows a construction site. "
-                       "If it does, identify any safety issues or hazards, categorize them, and provide a detailed description, "
-                       "and suggest steps to resolve them. If it's not a construction site, simply state that")
+        file_type = file.name.split('.')[-1].lower()
+        if file_type in ['jpg', 'jpeg', 'png', 'bmp']:
+            return analyze_image(file)
+        elif file_type in ['mp4', 'avi', 'mov', 'webm']:
+            return analyze_video(file)
+        else:
+            return "Unsupported file type. Please upload an image or video file."
+    except Exception as e:
+        logger.error(f"Error analyzing file: {str(e)}")
+        return f"Error analyzing file: {str(e)}"
 
-        for i, file in enumerate(input_files):
-            file_type = file.name.split('.')[-1].lower()
-            if file_type in ['jpg', 'jpeg', 'png', 'bmp']:
-                # Process image
-                image = Image.open(file.name)
-                resized_image = resize_image(image)
-                image_data_url = f"data:image/png;base64,{encode_image(resized_image)}"
-                content_type = "image"
-            elif file_type in ['mp4', 'avi', 'mov', 'webm']:
-                # Process video
-                frames = extract_frames_from_video(file.name)
-                image_data_url = f"data:image/png;base64,{encode_image(Image.fromarray(frames[0]))}"  # Use the first frame
-                content_type = "video"
-            else:
-                results.append((f"File {i+1} analysis", f"Unsupported file type: {file_type}"))
-                continue
+def analyze_image(image_file):
+    image = Image.open(image_file.name)
+    resized_image = resize_image(image)
+    image_data_url = f"data:image/png;base64,{encode_image(resized_image)}"
+    
+    instruction = ("You are an AI assistant specialized in analyzing images for safety issues. "
+                   "Your task is first to explain what you see in the image and determine if the image shows a construction site. "
+                   "If it does, identify any safety issues or hazards, categorize them, and provide a detailed description, "
+                   "and suggest steps to resolve them. If it's not a construction site, simply state that")
 
-            messages = [
+    messages = [
+        {
+            "role": "user",
+            "content": [
                 {
-                    "role": "user",
-                    "content": [
-                        {
-                            "type": "text",
-                            "text": f"{instruction}\n\nAnalyze this {content_type} (File {i+1}/{len(input_files)}). First, determine if it's a construction site. If it is, explain the {content_type} in detail, focusing on safety aspects. If it's not, briefly describe what you see."
-                        },
-                        {
-                            "type": "image_url",
-                            "image_url": {
-                                "url": image_data_url
-                            }
-                        }
-                    ]
+                    "type": "text",
+                    "text": f"{instruction}\n\nAnalyze this image. First, determine if it's a construction site. If it is, explain the image in detail, focusing on safety aspects. If it's not, briefly describe what you see."
+                },
+                {
+                    "type": "image_url",
+                    "image_url": {
+                        "url": image_data_url
+                    }
                 }
             ]
-            
-            completion = client.chat.completions.create(
-                model="llama-3.2-90b-vision-preview",
-                messages=messages,
-                temperature=0.7,
-                max_tokens=1000,
-                top_p=1,
-                stream=False,
-                stop=None
-            )
-            result = completion.choices[0].message.content
-            results.append((f"File {i+1} analysis ({content_type})", result))
+        }
+    ]
+    
+    completion = client.chat.completions.create(
+        model="llama-3.2-90b-vision-preview",
+        messages=messages,
+        temperature=0.7,
+        max_tokens=1000,
+        top_p=1,
+        stream=False,
+        stop=None
+    )
+    
+    return completion.choices[0].message.content
+
+def analyze_video(video_file):
+    frames = extract_frames_from_video(video_file.name)
+    results = []
+    
+    instruction = ("You are an AI assistant specialized in analyzing images for safety issues. "
+                   "Your task is first to explain what you see in the image and determine if the image shows a construction site. "
+                   "If it does, identify any safety issues or hazards, categorize them, and provide a detailed description, "
+                   "and suggest steps to resolve them. If it's not a construction site, simply state that")
 
-            # If it's a video, analyze additional frames
-            if content_type == "video" and len(frames) > 1:
-                for j, frame in enumerate(frames[1:], start=2):
-                    frame_data_url = f"data:image/png;base64,{encode_image(Image.fromarray(frame))}"
-                    messages = [
-                        {
-                            "role": "user",
-                            "content": [
-                                {
-                                    "type": "text",
-                                    "text": f"{instruction}\n\nAnalyze this additional frame from the video (File {i+1}, Frame {j}/{len(frames)}). Focus on any new or changed safety aspects compared to the previous frame."
-                                },
-                                {
-                                    "type": "image_url",
-                                    "image_url": {
-                                        "url": frame_data_url
-                                    }
-                                }
-                            ]
+    for i, frame in enumerate(frames):
+        image_data_url = f"data:image/png;base64,{encode_image(frame)}"
+        messages = [
+            {
+                "role": "user",
+                "content": [
+                    {
+                        "type": "text",
+                        "text": f"{instruction}\n\nAnalyze this frame from a video (Frame {i+1}/{len(frames)}). First, explain the video and then determine if it's a construction site. If it is, explain what you observe, focusing on safety aspects. If it's not, briefly describe what you see."
+                    },
+                    {
+                        "type": "image_url",
+                        "image_url": {
+                            "url": image_data_url
                         }
-                    ]
-                    completion = client.chat.completions.create(
-                        model="llama-3.2-90b-vision-preview",
-                        messages=messages,
-                        temperature=0.7,
-                        max_tokens=1000,
-                        top_p=1,
-                        stream=False,
-                        stop=None
-                    )
-                    result = completion.choices[0].message.content
-                    results.append((f"File {i+1} analysis (video frame {j})", result))
+                    }
+                ]
+            }
+        ]
+        completion = client.chat.completions.create(
+            model="llama-3.2-90b-vision-preview",
+            messages=messages,
+            temperature=0.7,
+            max_tokens=1000,
+            top_p=1,
+            stream=False,
+            stop=None
+        )
+        results.append(f"Frame {i+1} analysis:\n{completion.choices[0].message.content}\n\n")
+    
+    return "\n".join(results)
 
-        logger.info("Analysis completed successfully")
-        return results
-    except Exception as e:
-        logger.error(f"Error during analysis: {str(e)}")
-        logger.error(traceback.format_exc())
-        return [("Analysis error", f"Error during analysis: {str(e)}")]
 
         
 def chat_about_image(message, chat_history):
@@ -281,24 +276,19 @@ with gr.Blocks(css=custom_css, theme=gr.themes.Soft()) as iface:
         """
     )
     
-    # Single upload window for both images and videos
+    # First row: Combined file upload for images and videos
     with gr.Row():
-        input_files = gr.File(
-            label="Upload Construction Site Images or Videos",
-            file_count="multiple",
-            type="filepath",  # Changed from "file" to "filepath"
-            elem_classes="image-container"
-        )
+        file_input = gr.File(label="Upload Construction Site Images or Videos", file_count="multiple", type="file", elem_classes="file-container")
     
-    # Analyze Safety Hazards Button
+    # Second row: Analyze Safety Hazards Button
     with gr.Row():
         analyze_button = gr.Button("🔍 Analyze Safety Hazards", elem_classes="analyze-button")
     
-    # Chat Interface (Safety Analysis Results)
+    # Third row: Chat Interface (Safety Analysis Results)
     with gr.Row():
         chatbot = gr.Chatbot(label="Safety Analysis Results and Expert Chat", elem_classes="chatbot")
     
-    # Question Bar
+    # Fourth row: Question Bar
     with gr.Row():
         msg = gr.Textbox(
             label="Ask about safety measures or regulations",
@@ -307,14 +297,21 @@ with gr.Blocks(css=custom_css, theme=gr.themes.Soft()) as iface:
             elem_classes="chat-input"
         )
 
-    # Clear Chat and Download Report Buttons
+    # Fifth row: Clear Chat and Download Report Buttons
     with gr.Row():
         clear = gr.Button("🗑️ Clear Chat", elem_classes="clear-button")
         download_button = gr.Button("📥 Download Report", elem_classes="download-button")
 
-    # File component to handle the download
+    # Add a file component to handle the download
     report_file = gr.File(label="Download Safety Analysis Report")
 
+    def process_files(files):
+        results = []
+        for file in files:
+            result = analyze_file(file)
+            results.append((file.name, result))
+        return results
+
     def update_chat(history, new_messages):
         history = history or []
         for title, content in new_messages:
@@ -322,8 +319,8 @@ with gr.Blocks(css=custom_css, theme=gr.themes.Soft()) as iface:
         return history
 
     analyze_button.click(
-        analyze_mixed_input,
-        inputs=[input_files],
+        process_files,
+        inputs=[file_input],
         outputs=[chatbot],
         postprocess=lambda x: update_chat(chatbot.value, x)
     )
@@ -343,10 +340,11 @@ with gr.Blocks(css=custom_css, theme=gr.themes.Soft()) as iface:
         """
     )
 
+# Launch the app
 if __name__ == "__main__":
     try:
         iface.launch(debug=True)
     except Exception as e:
         logger.error(f"Error when trying to launch the interface: {str(e)}")
         logger.error(traceback.format_exc())
-        print("Failed to launch the Gradio interface. Please check the logs for more information.")
+        print("Failed to launch the Gradio interface. Please check the logs for more information.")