Update app.py

app.py

@@ -1,36 +1,37 @@
-import streamlit as st
+import streamlit as st 
 from phi.agent import Agent
 from phi.model.google import Gemini
 from phi.tools.duckduckgo import DuckDuckGo
-from google.generativeai import upload_file, get_file
+from google.generativeai import upload_file,get_file
 import google.generativeai as genai
 
-from moviepy.editor import VideoFileClip
 import time
 from pathlib import Path
+
 import tempfile
+
 from dotenv import load_dotenv
+load_dotenv()
+
 import os
 
-# Load API key from environment
-load_dotenv()
-API_KEY = os.getenv("GOOGLE_API_KEY")
+API_KEY=os.getenv("GOOGLE_API_KEY")
 if API_KEY:
     genai.configure(api_key=API_KEY)
 
-# Streamlit Page Configuration
+# Page configuration
 st.set_page_config(
-    page_title="Multimodal AI Agent - Video Summarizer",
+    page_title="Multimodal AI Agent- Video Summarizer",
     page_icon="🎥",
-    layout="wide",
+    layout="wide"
 )
 
 st.title("Phidata Video AI Summarizer Agent 🎥🎤🖬")
 st.header("Powered by Gemini 2.0 Flash Exp")
 
+
 @st.cache_resource
 def initialize_agent():
-    """Initialize the multimodal AI agent."""
     return Agent(
         name="Video AI Summarizer",
         model=Gemini(id="gemini-2.0-flash-exp"),
@@ -38,8 +39,8 @@ def initialize_agent():
         markdown=True,
     )
 
-# Initialize the agent
-multimodal_agent = initialize_agent()
+## Initialize the agent
+multimodal_Agent=initialize_agent()
 
 # File uploader
 video_file = st.file_uploader(
@@ -47,39 +48,31 @@ video_file = st.file_uploader(
 )
 
 if video_file:
-    # Temporary storage for uploaded video
     with tempfile.NamedTemporaryFile(delete=False, suffix='.mp4') as temp_video:
         temp_video.write(video_file.read())
         video_path = temp_video.name
 
-    # Display the uploaded video
     st.video(video_path, format="video/mp4", start_time=0)
 
-    # Display video duration
-    clip = VideoFileClip(video_path)
-    st.info(f"Video Duration: {clip.duration:.2f} seconds")
-
-    # Text area for user queries
     user_query = st.text_area(
         "What insights are you seeking from the video?",
         placeholder="Ask anything about the video content. The AI agent will analyze and gather additional context if needed.",
         help="Provide specific questions or insights you want from the video."
     )
 
-    # Analysis button
     if st.button("🔍 Analyze Video", key="analyze_video_button"):
         if not user_query:
             st.warning("Please enter a question or insight to analyze the video.")
         else:
             try:
                 with st.spinner("Processing video and gathering insights..."):
-                    # Upload and process the video file
+                    # Upload and process video file
                     processed_video = upload_file(video_path)
                     while processed_video.state.name == "PROCESSING":
                         time.sleep(1)
                         processed_video = get_file(processed_video.name)
 
-                    # Generate a prompt for AI analysis
+                    # Prompt generation for analysis
                     analysis_prompt = (
                         f"""
                         Analyze the uploaded video for content and context.
@@ -90,8 +83,8 @@ if video_file:
                         """
                     )
 
-                    # AI agent processes the query
-                    response = multimodal_agent.run(analysis_prompt, videos=[processed_video])
+                    # AI agent processing
+                    response = multimodal_Agent.run(analysis_prompt, videos=[processed_video])
 
                 # Display the result
                 st.subheader("Analysis Result")
@@ -102,40 +95,19 @@ if video_file:
             finally:
                 # Clean up temporary video file
                 Path(video_path).unlink(missing_ok=True)
-
-    # Additional features
-    if st.checkbox("Show Video Frames", help="Preview frames from the video."):
-        frame_rate = st.slider("Select Frame Rate (frames per second)", 1, 10, 1)
-        for i in range(0, int(clip.duration), frame_rate):
-            frame = clip.get_frame(i)
-            st.image(frame, caption=f"Frame at {i} seconds")
-
-    if st.checkbox("Show Video Transcription", help="Generate a transcription of the video's audio."):
-        try:
-            import whisper
-            model = whisper.load_model("base")
-            transcription = model.transcribe(video_path)
-            st.text_area("Video Transcription", transcription['text'], height=300)
-        except ImportError:
-            st.error("The 'whisper' library is not installed. Please install it to enable transcription.")
-
 else:
     st.info("Upload a video file to begin analysis.")
 
-# Style adjustments
+# Customize text area height
 st.markdown(
     """
     <style>
     .stTextArea textarea {
         height: 100px;
     }
-    .stButton > button {
-        background-color: #007ACC;
-        color: white;
-        border-radius: 5px;
-    }
     </style>
     """,
     unsafe_allow_html=True
 )
 
+