diff --git a/.gitignore b/.gitignore
index 7386e39..1ce164e 100644
--- a/.gitignore
+++ b/.gitignore
@@ -301,4 +301,7 @@ dist
 .pnp.*
 
 # Yolo
-*.pt
\ No newline at end of file
+*.pt
+
+# Downloads
+downloaded_videos/
diff --git a/jd-clone/index.json b/jd-clone/index.json
new file mode 100644
index 0000000..9570087
--- /dev/null
+++ b/jd-clone/index.json
@@ -0,0 +1,18 @@
+{
+    "songs": {
+        "Butter": {
+            "Name": "Butter",
+            "Artist": "BTS",
+            "Album": "Butter",
+            "Year": 2021,
+            "Genre": "K-Pop",
+            "GameData": {
+                "moves": "https://cdn.ovosimpatico.com/jdlo/maps/Butter/poses.json",
+                "video": "https://cdn.ovosimpatico.com/jdlo/maps/Butter/Butter_ULTRA.webm",
+                "audio": "https://cdn.ovosimpatico.com/jdlo/maps/Butter/Butter.ogg",
+                "cover": "https://cdn.ovosimpatico.com/jdlo/maps/Butter/butter_cover_generic.png",
+                "audio_preview": "https://cdn.ovosimpatico.com/jdlo/maps/Butter/butter_audiopreview.ogg"
+            }
+        }
+    }
+}
diff --git a/jd-clone/package.json b/jd-clone/package.json
index ba7a2dd..20973d1 100644
--- a/jd-clone/package.json
+++ b/jd-clone/package.json
@@ -5,7 +5,7 @@
   "type": "module",
   "scripts": {
     "dev": "vite",
-    "build": "tsc -b && vite build",
+    "build": "tsc -b && vite build && cp ../index.json dist/",
     "lint": "eslint .",
     "preview": "vite preview"
   },
diff --git a/jd-clone/src/components/common/AudioPreview.tsx b/jd-clone/src/components/common/AudioPreview.tsx
new file mode 100644
index 0000000..28aeafb
--- /dev/null
+++ b/jd-clone/src/components/common/AudioPreview.tsx
@@ -0,0 +1,61 @@
+import { useEffect, useRef, useState } from 'react';
+import useAppStore from '../../store/app-store';
+
+interface AudioPreviewProps {
+  src: string;
+  autoPlay?: boolean;
+}
+
+const AudioPreview = ({ src, autoPlay = false }: AudioPreviewProps) => {
+  const audioRef = useRef<HTMLAudioElement | null>(null);
+  const [isPlaying, setIsPlaying] = useState(false);
+  const masterVolume = useAppStore(state => state.settings.volume.master);
+  const musicVolume = useAppStore(state => state.settings.volume.music);
+
+  useEffect(() => {
+    if (!audioRef.current) return;
+
+    // Set volume based on app settings
+    const volume = masterVolume * musicVolume;
+    audioRef.current.volume = volume;
+
+    // Autoplay if needed
+    if (autoPlay && audioRef.current) {
+      audioRef.current.play().catch(error => {
+        console.warn('Autoplay prevented:', error);
+      });
+    }
+
+    return () => {
+      if (audioRef.current) {
+        audioRef.current.pause();
+        audioRef.current.currentTime = 0;
+      }
+    };
+  }, [src, autoPlay, masterVolume, musicVolume]);
+
+  useEffect(() => {
+    const audio = audioRef.current;
+    if (!audio) return;
+
+    const handlePlay = () => setIsPlaying(true);
+    const handlePause = () => setIsPlaying(false);
+    const handleEnded = () => setIsPlaying(false);
+
+    audio.addEventListener('play', handlePlay);
+    audio.addEventListener('pause', handlePause);
+    audio.addEventListener('ended', handleEnded);
+
+    return () => {
+      audio.removeEventListener('play', handlePlay);
+      audio.removeEventListener('pause', handlePause);
+      audio.removeEventListener('ended', handleEnded);
+    };
+  }, []);
+
+  return (
+    <audio ref={audioRef} src={src} loop={false} preload="auto" />
+  );
+};
+
+export default AudioPreview;
\ No newline at end of file
diff --git a/jd-clone/src/components/game/VideoPlayer.scss b/jd-clone/src/components/game/VideoPlayer.scss
new file mode 100644
index 0000000..32c05c4
--- /dev/null
+++ b/jd-clone/src/components/game/VideoPlayer.scss
@@ -0,0 +1,14 @@
+.video-player {
+  position: relative;
+  overflow: hidden;
+  background-color: black;
+  border-radius: 8px;
+  box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
+
+  video {
+    display: block;
+    object-fit: contain;
+    width: 100%;
+    height: 100%;
+  }
+}
\ No newline at end of file
diff --git a/jd-clone/src/components/game/VideoPlayer.tsx b/jd-clone/src/components/game/VideoPlayer.tsx
new file mode 100644
index 0000000..d8d8550
--- /dev/null
+++ b/jd-clone/src/components/game/VideoPlayer.tsx
@@ -0,0 +1,168 @@
+import { useEffect, useRef } from 'react';
+import useAppStore from '../../store/app-store';
+import './VideoPlayer.scss';
+
+interface VideoPlayerProps {
+  src: string;
+  audioSrc?: string;
+  playing: boolean;
+  muted?: boolean;
+  width?: number | string;
+  height?: number | string;
+  onEnded?: () => void;
+}
+
+const VideoPlayer = ({
+  src,
+  audioSrc,
+  playing,
+  muted = false,
+  width = '100%',
+  height = '100%',
+  onEnded
+}: VideoPlayerProps) => {
+  const videoRef = useRef<HTMLVideoElement | null>(null);
+  const audioRef = useRef<HTMLAudioElement | null>(null);
+  const masterVolume = useAppStore(state => state.settings.volume.master);
+  const musicVolume = useAppStore(state => state.settings.volume.music);
+
+  // Manage play/pause state for video
+  useEffect(() => {
+    const video = videoRef.current;
+    if (!video) return;
+
+    if (playing) {
+      video.play().catch(error => {
+        console.warn('Video playback prevented:', error);
+      });
+    } else {
+      video.pause();
+    }
+  }, [playing]);
+
+  // Manage play/pause state for audio
+  useEffect(() => {
+    const audio = audioRef.current;
+    if (!audio || !audioSrc) return;
+
+    if (playing) {
+      // Sync with video if needed
+      if (videoRef.current) {
+        audio.currentTime = videoRef.current.currentTime;
+      }
+
+      audio.play().catch(error => {
+        console.warn('Audio playback prevented:', error);
+      });
+    } else {
+      audio.pause();
+    }
+  }, [playing, audioSrc]);
+
+  // Sync audio with video when video seeks or loads
+  useEffect(() => {
+    const video = videoRef.current;
+    const audio = audioRef.current;
+    if (!video || !audio || !audioSrc) return;
+
+    const handleTimeUpdate = () => {
+      // Only sync if the difference is significant (more than 0.1 seconds)
+      if (Math.abs(video.currentTime - audio.currentTime) > 0.1) {
+        audio.currentTime = video.currentTime;
+      }
+    };
+
+    const handlePlay = () => {
+      if (playing) {
+        audio.play().catch(error => {
+          console.warn('Audio playback prevented:', error);
+        });
+      }
+    };
+
+    const handlePause = () => {
+      audio.pause();
+    };
+
+    video.addEventListener('seeked', handleTimeUpdate);
+    video.addEventListener('play', handlePlay);
+    video.addEventListener('pause', handlePause);
+
+    return () => {
+      video.removeEventListener('seeked', handleTimeUpdate);
+      video.removeEventListener('play', handlePlay);
+      video.removeEventListener('pause', handlePause);
+    };
+  }, [playing, audioSrc]);
+
+  // Manage volume for video
+  useEffect(() => {
+    const video = videoRef.current;
+    if (!video) return;
+
+    // If we have a separate audio source, mute the video
+    if (audioSrc) {
+      video.muted = true;
+    } else {
+      // Otherwise use the video's audio
+      if (muted) {
+        video.muted = true;
+      } else {
+        video.muted = false;
+        video.volume = masterVolume * musicVolume;
+      }
+    }
+  }, [masterVolume, musicVolume, muted, audioSrc]);
+
+  // Manage volume for audio
+  useEffect(() => {
+    const audio = audioRef.current;
+    if (!audio || !audioSrc) return;
+
+    if (muted) {
+      audio.muted = true;
+    } else {
+      audio.muted = false;
+      audio.volume = masterVolume * musicVolume;
+    }
+  }, [masterVolume, musicVolume, muted, audioSrc]);
+
+  // Setup event handlers for video
+  useEffect(() => {
+    const video = videoRef.current;
+    if (!video) return;
+
+    const handleEnded = () => {
+      if (onEnded) onEnded();
+    };
+
+    video.addEventListener('ended', handleEnded);
+
+    return () => {
+      video.removeEventListener('ended', handleEnded);
+    };
+  }, [onEnded]);
+
+  return (
+    <div className="video-player" style={{ width, height }}>
+      <video
+        ref={videoRef}
+        src={src}
+        preload="auto"
+        playsInline
+        width="100%"
+        height="100%"
+        muted={!!audioSrc || muted} // Always mute video if using separate audio
+      />
+      {audioSrc && (
+        <audio
+          ref={audioRef}
+          src={audioSrc}
+          preload="auto"
+        />
+      )}
+    </div>
+  );
+};
+
+export default VideoPlayer;
\ No newline at end of file
diff --git a/jd-clone/src/pages/gameplay/GameplayPage.scss b/jd-clone/src/pages/gameplay/GameplayPage.scss
index 71f298c..04a00d2 100644
--- a/jd-clone/src/pages/gameplay/GameplayPage.scss
+++ b/jd-clone/src/pages/gameplay/GameplayPage.scss
@@ -249,6 +249,47 @@
     align-items: center;
     gap: 1rem;
   }
+
+  // Completion overlay
+  &__completion-overlay {
+    position: fixed;
+    top: 0;
+    left: 0;
+    right: 0;
+    bottom: 0;
+    background-color: rgba(0, 0, 0, 0.7);
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    z-index: 100;
+    animation: fade-in 0.5s ease-out;
+  }
+
+  &__completion-content {
+    background-color: rgba(26, 26, 26, 0.9);
+    border-radius: 12px;
+    padding: 3rem;
+    text-align: center;
+    box-shadow: 0 0 30px rgba(255, 215, 0, 0.3);
+    animation: scale-in 0.5s ease-out;
+
+    h2 {
+      font-size: 3.5rem;
+      margin-top: 0;
+      margin-bottom: 1.5rem;
+      color: #ffca3a;
+    }
+
+    p {
+      font-size: 1.8rem;
+      margin-bottom: 1rem;
+
+      &:last-child {
+        color: #8ac926;
+        font-weight: bold;
+      }
+    }
+  }
 }
 
 @keyframes feedback-pulse {
@@ -263,4 +304,24 @@
     transform: scale(1);
     opacity: 1;
   }
+}
+
+@keyframes fade-in {
+  from {
+    opacity: 0;
+  }
+  to {
+    opacity: 1;
+  }
+}
+
+@keyframes scale-in {
+  from {
+    transform: scale(0.8);
+    opacity: 0;
+  }
+  to {
+    transform: scale(1);
+    opacity: 1;
+  }
 }
\ No newline at end of file
diff --git a/jd-clone/src/pages/gameplay/GameplayPage.tsx b/jd-clone/src/pages/gameplay/GameplayPage.tsx
index 2764ca6..fe589bc 100644
--- a/jd-clone/src/pages/gameplay/GameplayPage.tsx
+++ b/jd-clone/src/pages/gameplay/GameplayPage.tsx
@@ -1,6 +1,7 @@
 import { useEffect, useState } from 'react';
 import { useNavigate } from 'react-router-dom';
 import { PoseRenderer } from '../../components/game/PoseRenderer';
+import VideoPlayer from '../../components/game/VideoPlayer';
 import usePoseDetection from '../../hooks/usePoseDetection';
 import useInputDetection from '../../hooks/useInputDetection';
 import useControllerDetection from '../../hooks/useControllerDetection';
@@ -19,6 +20,7 @@ function GameplayPage() {
   const [gameStarted, setGameStarted] = useState(false);
   const [countdown, setCountdown] = useState(3);
   const [error, setError] = useState<string | null>(null);
+  const [gameCompleted, setGameCompleted] = useState(false);
 
   // Get app state
   const selectedSongId = useAppStore(state => state.selectedSongId);
@@ -101,6 +103,7 @@ function GameplayPage() {
 
     // Reset game state
     resetGameState();
+    setGameCompleted(false);
 
     loadGameData();
 
@@ -142,6 +145,16 @@ function GameplayPage() {
     navigate('/results');
   };
 
+  // Handle video ended
+  const handleVideoEnded = () => {
+    setGameCompleted(true);
+    endGame();
+    // Wait a short moment before navigating to results
+    setTimeout(() => {
+      navigate('/results');
+    }, 1500);
+  };
+
   // If we don't have a selected song, redirect to setup
   if (!selectedSongId) {
     useEffect(() => {
@@ -192,6 +205,21 @@ function GameplayPage() {
     );
   };
 
+  // Helper to render game completion overlay
+  const renderCompletionOverlay = () => {
+    if (!gameCompleted) return null;
+
+    return (
+      <div className="gameplay-page__completion-overlay">
+        <div className="gameplay-page__completion-content">
+          <h2>Song Complete!</h2>
+          <p>Great job dancing!</p>
+          <p>Redirecting to results...</p>
+        </div>
+      </div>
+    );
+  };
+
   return (
     <div className="gameplay-page gameplay-page--tv-mode">
       {loading ? (
@@ -271,11 +299,19 @@ function GameplayPage() {
 
           <div className="gameplay-page__content">
             <div className="gameplay-page__video-container">
-              {/* Video will go here */}
-              <div className="gameplay-page__video-placeholder">
-                <p>Dance video will play here</p>
-                <p>Song: {song?.title} by {song?.artist}</p>
-              </div>
+              {song && song.GameData ? (
+                <VideoPlayer
+                  src={song.GameData.video}
+                  audioSrc={song.GameData.audio}
+                  playing={isPlaying && !isPaused}
+                  onEnded={handleVideoEnded}
+                />
+              ) : (
+                <div className="gameplay-page__video-placeholder">
+                  <p>Dance video will play here</p>
+                  <p>Song information not available</p>
+                </div>
+              )}
             </div>
 
             <div className="gameplay-page__camera-container">
@@ -302,6 +338,7 @@ function GameplayPage() {
           </div>
 
           {renderPauseOverlay()}
+          {renderCompletionOverlay()}
         </>
       )}
     </div>
diff --git a/jd-clone/src/pages/results/ResultsPage.tsx b/jd-clone/src/pages/results/ResultsPage.tsx
index 0fca763..35056be 100644
--- a/jd-clone/src/pages/results/ResultsPage.tsx
+++ b/jd-clone/src/pages/results/ResultsPage.tsx
@@ -46,8 +46,8 @@ function ResultsPage() {
         <h1 className="results-page__title">Results</h1>
 
         <div className="results-page__song-info">
-          <h2>{selectedSong?.title || 'Unknown Song'}</h2>
-          <p>{selectedSong?.artist || 'Unknown Artist'}</p>
+          <h2>{selectedSong?.Name || 'Unknown Song'}</h2>
+          <p>{selectedSong?.Artist || 'Unknown Artist'}</p>
         </div>
 
         <div className="results-page__score-container">
diff --git a/jd-clone/src/pages/setup/GameSetupPage.tsx b/jd-clone/src/pages/setup/GameSetupPage.tsx
index 3724017..0778a05 100644
--- a/jd-clone/src/pages/setup/GameSetupPage.tsx
+++ b/jd-clone/src/pages/setup/GameSetupPage.tsx
@@ -2,6 +2,7 @@ import React, { useEffect, useState } from 'react';
 import { useNavigate } from 'react-router-dom';
 import { Button } from '../../components/common/Button';
 import { PoseRenderer } from '../../components/game/PoseRenderer';
+import AudioPreview from '../../components/common/AudioPreview';
 import usePoseDetection from '../../hooks/usePoseDetection';
 import useAppStore from '../../store/app-store';
 import songService from '../../services/song-service';
@@ -13,6 +14,7 @@ function GameSetupPage() {
   const [songs, setSongs] = useState<Song[]>([]);
   const [loading, setLoading] = useState(true);
   const [selectedSongId, setSelectedSongId] = useState<string | null>(null);
+  const [currentPreviewUrl, setCurrentPreviewUrl] = useState<string | null>(null);
 
   const difficulty = useAppStore(state => state.settings.difficulty);
   const setDifficulty = useAppStore(state => state.setDifficulty);
@@ -31,6 +33,8 @@ function GameSetupPage() {
         // Select the first song by default
         if (songList.length > 0 && !selectedSongId) {
           setSelectedSongId(songList[0].id);
+          // Set the preview URL for the first song
+          setCurrentPreviewUrl(songList[0].GameData.audio_preview);
         }
       } catch (error) {
         console.error('Failed to load songs:', error);
@@ -48,6 +52,12 @@ function GameSetupPage() {
 
   const handleSongSelect = (songId: string) => {
     setSelectedSongId(songId);
+
+    // Update the audio preview URL when selecting a new song
+    const selectedSong = songs.find(song => song.id === songId);
+    if (selectedSong) {
+      setCurrentPreviewUrl(selectedSong.GameData.audio_preview);
+    }
   };
 
   const handleStartGame = () => {
@@ -103,27 +113,32 @@ function GameSetupPage() {
                   className={`setup-page__song-item ${selectedSongId === song.id ? 'setup-page__song-item--selected' : ''}`}
                   onClick={() => handleSongSelect(song.id)}
                 >
-                  <div className="setup-page__song-cover" style={{ backgroundImage: `url(${song.coverUrl})` }} />
+                  <div className="setup-page__song-cover" style={{ backgroundImage: `url(${song.GameData.cover})` }} />
                   <div className="setup-page__song-info">
-                    <h3 className="setup-page__song-title">{song.title}</h3>
-                    <p className="setup-page__song-artist">{song.artist}</p>
-                    <p className="setup-page__song-duration">{Math.floor(song.duration / 60)}:{(song.duration % 60).toString().padStart(2, '0')}</p>
+                    <h3 className="setup-page__song-title">{song.Name}</h3>
+                    <p className="setup-page__song-artist">{song.Artist}</p>
+                    <p className="setup-page__song-genre">{song.Genre || 'Unknown genre'}</p>
                   </div>
                 </div>
               ))}
             </div>
           )}
+
+          {/* Audio Preview Player */}
+          {currentPreviewUrl && (
+            <AudioPreview src={currentPreviewUrl} autoPlay={true} />
+          )}
         </div>
 
         <div className="setup-page__difficulty-selection">
           <h2>Select Difficulty</h2>
           <div className="setup-page__difficulty-buttons">
-            {selectedSong?.difficulty.map((diff) => (
+            {(['easy', 'medium', 'hard', 'extreme'] as DifficultyLevel[]).map((diff) => (
               <button
                 key={diff}
                 className={`setup-page__difficulty-button setup-page__difficulty-button--${diff} ${difficulty === diff ? 'setup-page__difficulty-button--selected' : ''}`}
                 onClick={() => handleDifficultyChange(diff)}
-                disabled={!selectedSong.difficulty.includes(diff)}
+                disabled={selectedSong?.difficulty ? !selectedSong.difficulty.includes(diff) : false}
               >
                 {diff.toUpperCase()}
               </button>
diff --git a/jd-clone/src/services/song-service.ts b/jd-clone/src/services/song-service.ts
index adcf14b..6ae8d06 100644
--- a/jd-clone/src/services/song-service.ts
+++ b/jd-clone/src/services/song-service.ts
@@ -1,140 +1,101 @@
 import { Song, Choreography, DifficultyLevel, Move } from '../types';
 
-// Mock song data for development purposes
-const MOCK_SONGS: Song[] = [
-  {
-    id: 'song1',
-    title: 'Dance The Night',
-    artist: 'Dua Lipa',
-    bpm: 120,
-    duration: 176,
-    coverUrl: 'https://example.com/covers/dance-the-night.jpg',
-    audioUrl: 'https://example.com/songs/dance-the-night.mp3',
-    videoUrl: 'https://example.com/videos/dance-the-night.mp4',
-    difficulty: ['easy', 'medium', 'hard'],
-    tags: ['pop', 'upbeat', 'disco']
-  },
-  {
-    id: 'song2',
-    title: 'Levitating',
-    artist: 'Dua Lipa ft. DaBaby',
-    bpm: 103,
-    duration: 203,
-    coverUrl: 'https://example.com/covers/levitating.jpg',
-    audioUrl: 'https://example.com/songs/levitating.mp3',
-    videoUrl: 'https://example.com/videos/levitating.mp4',
-    difficulty: ['easy', 'medium', 'hard', 'extreme'],
-    tags: ['pop', 'upbeat', 'disco']
-  },
-  {
-    id: 'song3',
-    title: 'Physical',
-    artist: 'Dua Lipa',
-    bpm: 124,
-    duration: 183,
-    coverUrl: 'https://example.com/covers/physical.jpg',
-    audioUrl: 'https://example.com/songs/physical.mp3',
-    videoUrl: 'https://example.com/videos/physical.mp4',
-    difficulty: ['medium', 'hard', 'extreme'],
-    tags: ['pop', 'dance', 'workout']
-  }
-];
-
-// Mock choreography data with placeholder moves
-const MOCK_CHOREOGRAPHIES: Record<string, Record<DifficultyLevel, Choreography>> = {
-  song1: {
-    easy: {
-      songId: 'song1',
-      difficulty: 'easy',
-      moves: Array(20).fill(null).map((_, index) => ({
-        id: `song1-easy-move-${index}`,
-        startTime: index * 8000,
-        duration: 4000,
-        keyPosePoints: [], // This would contain actual pose landmarks
-        difficulty: 'easy',
-        score: 100
-      }))
-    },
-    medium: {
-      songId: 'song1',
-      difficulty: 'medium',
-      moves: Array(30).fill(null).map((_, index) => ({
-        id: `song1-medium-move-${index}`,
-        startTime: index * 6000,
-        duration: 3000,
-        keyPosePoints: [],
-        difficulty: 'medium',
-        score: 150
-      }))
-    },
-    hard: {
-      songId: 'song1',
-      difficulty: 'hard',
-      moves: Array(40).fill(null).map((_, index) => ({
-        id: `song1-hard-move-${index}`,
-        startTime: index * 4000,
-        duration: 2000,
-        keyPosePoints: [],
-        difficulty: 'hard',
-        score: 200
-      }))
-    },
-    extreme: {
-      songId: 'song1',
-      difficulty: 'extreme',
-      moves: Array(50).fill(null).map((_, index) => ({
-        id: `song1-extreme-move-${index}`,
-        startTime: index * 3000,
-        duration: 1500,
-        keyPosePoints: [],
-        difficulty: 'extreme',
-        score: 300
-      }))
-    }
-  }
-};
-
 class SongService {
   /**
    * Get all available songs
    */
   async getSongs(): Promise<Song[]> {
-    // In a real app, this would fetch from an API
-    return new Promise((resolve) => {
-      setTimeout(() => {
-        resolve(MOCK_SONGS);
-      }, 500);
-    });
+    try {
+      const response = await fetch('/index.json');
+      if (!response.ok) {
+        throw new Error(`Failed to fetch songs: ${response.statusText}`);
+      }
+
+      const data = await response.json();
+
+      // Transform the data to match our Song interface
+      return Object.entries(data.songs).map(([id, songData]: [string, any]) => ({
+        id,
+        ...songData,
+        // Add default difficulty levels since they're not in the JSON
+        difficulty: ['medium', 'hard']
+      }));
+    } catch (error) {
+      console.error('Error fetching songs:', error);
+      return [];
+    }
   }
 
   /**
    * Get a specific song by ID
    */
   async getSongById(id: string): Promise<Song | null> {
-    return new Promise((resolve) => {
-      setTimeout(() => {
-        const song = MOCK_SONGS.find(s => s.id === id) || null;
-        resolve(song);
-      }, 300);
-    });
+    try {
+      const songs = await this.getSongs();
+      return songs.find(song => song.id === id) || null;
+    } catch (error) {
+      console.error(`Error fetching song ${id}:`, error);
+      return null;
+    }
   }
 
   /**
    * Get choreography for a song at a specific difficulty
    */
   async getChoreography(songId: string, difficulty: DifficultyLevel): Promise<Choreography | null> {
-    return new Promise((resolve) => {
-      setTimeout(() => {
-        const songChoreographies = MOCK_CHOREOGRAPHIES[songId];
-        if (!songChoreographies) {
-          resolve(null);
-          return;
-        }
+    try {
+      const song = await this.getSongById(songId);
+      if (!song) {
+        return null;
+      }
 
-        const choreography = songChoreographies[difficulty];
-        resolve(choreography || null);
-      }, 500);
-    });
+      // Fetch the moves data from the URL in song.GameData.moves
+      const response = await fetch(song.GameData.moves);
+      if (!response.ok) {
+        throw new Error(`Failed to fetch choreography: ${response.statusText}`);
+      }
+
+      const choreographyData = await response.json();
+
+      // Process the choreography data based on difficulty
+      // This would need to be adjusted based on the actual format of the moves data
+      return {
+        songId,
+        difficulty,
+        moves: this.processChoreographyData(choreographyData, difficulty)
+      };
+    } catch (error) {
+      console.error(`Error fetching choreography for song ${songId}:`, error);
+      return null;
+    }
+  }
+
+  /**
+   * Process choreography data from the JSON file
+   * Note: This would need to be adjusted based on the actual data structure
+   */
+  private processChoreographyData(data: any, difficulty: DifficultyLevel): Move[] {
+    // This is a placeholder implementation
+    // You'll need to adapt this based on the actual format of your poses.json files
+    const moves: Move[] = [];
+
+    // Example implementation assuming data has a moves array
+    if (Array.isArray(data.moves)) {
+      data.moves.forEach((moveData: any, index: number) => {
+        moves.push({
+          id: `${difficulty}-move-${index}`,
+          startTime: moveData.startTime || index * 3000,
+          duration: moveData.duration || 2000,
+          keyPosePoints: moveData.keyPoints || [],
+          difficulty,
+          score: difficulty === 'easy' ? 100 :
+                 difficulty === 'medium' ? 150 :
+                 difficulty === 'hard' ? 200 : 300
+        });
+      });
+    }
+
+    return moves;
   }
 
   /**
@@ -144,12 +105,9 @@ class SongService {
     // This would use the pose detection API to analyze a video and generate choreography data
     console.log(`Generating choreography for ${videoUrl}, song ${songId}, difficulty ${difficulty}`);
 
-    // For now, just return a mock choreography
-    return new Promise((resolve) => {
-      setTimeout(() => {
-        resolve(MOCK_CHOREOGRAPHIES.song1[difficulty]);
-      }, 2000);
-    });
+    // This is a placeholder - in a real implementation,
+    // this would call an API to process the video and generate choreography
+    return null;
   }
 }
 
diff --git a/jd-clone/src/types/index.ts b/jd-clone/src/types/index.ts
index e8e2e46..d37e6d3 100644
--- a/jd-clone/src/types/index.ts
+++ b/jd-clone/src/types/index.ts
@@ -21,17 +21,23 @@ export interface PoseData {
 }
 
 // Game content types
+export interface SongGameData {
+  moves: string;
+  video: string;
+  audio: string;
+  cover: string;
+  audio_preview: string;
+}
+
 export interface Song {
   id: string;
-  title: string;
-  artist: string;
-  bpm: number;
-  duration: number;
-  coverUrl: string;
-  audioUrl: string;
-  videoUrl: string;
-  difficulty: DifficultyLevel[];
-  tags: string[];
+  Name: string;
+  Artist: string;
+  Album?: string;
+  Year?: number;
+  Genre?: string;
+  GameData: SongGameData;
+  difficulty?: DifficultyLevel[];
 }
 
 export type DifficultyLevel = 'easy' | 'medium' | 'hard' | 'extreme';
diff --git a/pose_detector_window.py b/pose_detector_window.py
index 3e54c25..2975637 100644
--- a/pose_detector_window.py
+++ b/pose_detector_window.py
@@ -1,5 +1,6 @@
 import argparse
 import json
+import math
 import os
 import time
 import urllib.request
@@ -77,7 +78,66 @@ def download_video(url: str, output_dir: str = "downloaded_videos") -> str:
     print(f"✅ Video downloaded successfully to {output_path}")
     return output_path
 
-def normalize_landmarks(landmarks: List[Dict], window_size: int = 5, poly_order: int = 4) -> List[Dict]:
+def normalize_landmarks_per_person(people_landmarks: List[Dict], window_size: int = 5, poly_order: int = 4) -> List[Dict]:
+    """Normalize landmarks over time for each person using Savitzky-Golay filter"""
+    if not people_landmarks:
+        return people_landmarks
+
+    # Reorganize by person ID
+    person_data = {}
+    for frame_data in people_landmarks:
+        frame_num = frame_data['frame']
+        timestamp = frame_data['timestamp']
+
+        for person in frame_data['people']:
+            person_id = person['person_id']
+
+            if person_id not in person_data:
+                person_data[person_id] = {
+                    'frames': [],
+                    'timestamps': [],
+                    'landmarks': []
+                }
+
+            person_data[person_id]['frames'].append(frame_num)
+            person_data[person_id]['timestamps'].append(timestamp)
+            person_data[person_id]['landmarks'].append(person['landmarks'])
+
+    # Normalize each person's landmarks
+    for person_id, data in person_data.items():
+        if len(data['landmarks']) >= window_size:
+            data['landmarks'] = normalize_landmarks(
+                data['landmarks'],
+                window_size=window_size,
+                poly_order=poly_order
+            )
+
+    # Reconstruct the frame data structure
+    normalized_data = []
+    for frame_data in people_landmarks:
+        frame_num = frame_data['frame']
+        timestamp = frame_data['timestamp']
+        new_people = []
+
+        for person in frame_data['people']:
+            person_id = person['person_id']
+            idx = person_data[person_id]['frames'].index(frame_num)
+
+            new_people.append({
+                'person_id': person_id,
+                'bbox': person['bbox'],
+                'landmarks': person_data[person_id]['landmarks'][idx]
+            })
+
+        normalized_data.append({
+            'frame': frame_num,
+            'timestamp': timestamp,
+            'people': new_people
+        })
+
+    return normalized_data
+
+def normalize_landmarks(landmarks: List[List[Dict]], window_size: int = 5, poly_order: int = 4) -> List[List[Dict]]:
     """Normalize landmarks over time using Savitzky-Golay filter to smooth motion"""
     if not landmarks or len(landmarks) < window_size:
         return landmarks
@@ -86,6 +146,12 @@ def normalize_landmarks(landmarks: List[Dict], window_size: int = 5, poly_order:
     if window_size % 2 == 0:
         window_size += 1
 
+    # Check if all frames have the same number of landmarks
+    if not all(len(frame) == len(landmarks[0]) for frame in landmarks):
+        # If inconsistent landmark counts, use a simpler approach (frame by frame smoothing)
+        print("⚠️ Warning: Inconsistent landmark counts across frames. Using simplified smoothing.")
+        return landmarks
+
     # Extract x, y values for each landmark
     landmark_count = len(landmarks[0])
     x_values = np.zeros((len(landmarks), landmark_count))
@@ -117,40 +183,226 @@ def normalize_landmarks(landmarks: List[Dict], window_size: int = 5, poly_order:
 
     return normalized_landmarks
 
+def calculate_iou(box1, box2):
+    """Calculate IoU (Intersection over Union) between two bounding boxes"""
+    # Extract coordinates
+    x1_1, y1_1, x2_1, y2_1 = box1
+    x1_2, y1_2, x2_2, y2_2 = box2
+
+    # Calculate intersection area
+    x_left = max(x1_1, x1_2)
+    y_top = max(y1_1, y1_2)
+    x_right = min(x2_1, x2_2)
+    y_bottom = min(y2_1, y2_2)
+
+    if x_right < x_left or y_bottom < y_top:
+        return 0.0
+
+    intersection_area = (x_right - x_left) * (y_bottom - y_top)
+
+    # Calculate union area
+    box1_area = (x2_1 - x1_1) * (y2_1 - y1_1)
+    box2_area = (x2_2 - x1_2) * (y2_2 - y1_2)
+    union_area = box1_area + box2_area - intersection_area
+
+    return intersection_area / union_area if union_area > 0 else 0
+
+def calculate_keypoint_distance(landmarks1, landmarks2):
+    """Calculate average distance between corresponding keypoints"""
+    if not landmarks1 or not landmarks2:
+        return float('inf')
+
+    # Create dictionary for fast lookup
+    kps1 = {lm['idx']: (lm['x'], lm['y']) for lm in landmarks1}
+    kps2 = {lm['idx']: (lm['x'], lm['y']) for lm in landmarks2}
+
+    # Find common keypoints
+    common_idx = set(kps1.keys()) & set(kps2.keys())
+    if not common_idx:
+        return float('inf')
+
+    # Calculate distance between corresponding keypoints
+    total_dist = 0
+    for idx in common_idx:
+        x1, y1 = kps1[idx]
+        x2, y2 = kps2[idx]
+        dist = math.sqrt((x1 - x2)**2 + (y1 - y2)**2)
+        total_dist += dist
+
+    return total_dist / len(common_idx)
+
+def assign_person_ids(current_people, previous_people, iou_threshold=0.3, distance_threshold=0.2):
+    """Assign stable IDs to people across frames based on IOU and keypoint distance"""
+    if not previous_people:
+        # First frame, assign new IDs to everyone
+        next_id = 0
+        for person in current_people:
+            person['person_id'] = next_id
+            next_id += 1
+        return current_people
+
+    # Create copy of current people to modify
+    assigned_people = []
+    unassigned_current = current_people.copy()
+
+    # Try to match current detections with previous ones
+    matched_prev_ids = set()
+
+    # Sort previous people by ID to maintain consistency in matching
+    sorted_prev = sorted(previous_people, key=lambda x: x['person_id'])
+
+    for prev_person in sorted_prev:
+        prev_id = prev_person['person_id']
+        prev_box = prev_person['bbox']
+        prev_landmarks = prev_person['landmarks']
+
+        best_match = None
+        best_score = float('inf')  # Lower is better for distance
+
+        for curr_person in unassigned_current:
+            curr_box = curr_person['bbox']
+            curr_landmarks = curr_person['landmarks']
+
+            # Calculate IoU between bounding boxes
+            iou = calculate_iou(prev_box, curr_box)
+
+            # Calculate keypoint distance
+            kp_dist = calculate_keypoint_distance(prev_landmarks, curr_landmarks)
+
+            # Combined score (lower is better)
+            score = kp_dist * (1.5 - iou)  # Favor high IoU and low distance
+
+            if (iou >= iou_threshold or kp_dist <= distance_threshold) and score < best_score:
+                best_match = curr_person
+                best_score = score
+
+        if best_match:
+            # Assign the previous ID to this person
+            best_match['person_id'] = prev_id
+            matched_prev_ids.add(prev_id)
+            assigned_people.append(best_match)
+            unassigned_current.remove(best_match)
+
+    # Find the next available ID
+    next_id = 0
+    existing_ids = {p['person_id'] for p in previous_people}
+    while next_id in existing_ids:
+        next_id += 1
+
+    # Assign new IDs to unmatched current detections
+    for person in unassigned_current:
+        person['person_id'] = next_id
+        assigned_people.append(person)
+        next_id += 1
+
+    return assigned_people
+
+def compress_pose_data(all_frame_data, frame_sampling=1, precision=3):
+    """Compress pose data to reduce JSON file size by reducing precision and sampling frames"""
+    compressed_data = []
+
+    # Process only every nth frame based on sampling rate
+    for i, frame_data in enumerate(all_frame_data):
+        if i % frame_sampling != 0:
+            continue
+
+        # Compress frame data
+        compressed_frame = {
+            'f': frame_data['frame'],  # Short key name
+            't': round(frame_data['timestamp'], 2),  # Reduce timestamp precision
+            'p': []  # Short key for people
+        }
+
+        # Process each person
+        for person in frame_data['people']:
+            # Only keep essential bbox info (we only need width/height for visualization)
+            x1, y1, x2, y2 = person['bbox']
+            width = x2 - x1
+            height = y2 - y1
+
+            compressed_person = {
+                'id': person['person_id'],  # Keep ID as is
+                'b': [round(x1, 1), round(y1, 1), round(width, 1), round(height, 1)],  # Simplified bbox with less precision
+                'k': []  # Short key for keypoints/landmarks
+            }
+
+            # Process each landmark with reduced precision
+            for lm in person['landmarks']:
+                compressed_person['k'].append([
+                    lm['idx'],  # Keep index as is (small integer)
+                    round(lm['x'], precision),  # Reduce coordinate precision
+                    round(lm['y'], precision),  # Reduce coordinate precision
+                    round(lm['confidence'], 2)  # Reduce confidence precision
+                ])
+
+            compressed_frame['p'].append(compressed_person)
+
+        compressed_data.append(compressed_frame)
+
+    return compressed_data
+
 def process_frame(frame: np.ndarray, model, detection_threshold: float = 0.5, show_preview: bool = False):
-    """Process a single frame with YOLOv11-pose"""
+    """Process a single frame with YOLOv11-pose, handling multiple people"""
     # Process with YOLO
     try:
         results = model.predict(frame, verbose=False, conf=detection_threshold)
 
         # Extract keypoints if available
-        landmarks_data = None
         processed_frame = None
+        people_data = []
 
         # Get frame dimensions
         h, w = frame.shape[:2]
 
         if results and len(results[0].keypoints.data) > 0:
-            # Get keypoints from the first detection
-            keypoints = results[0].keypoints.data[0]  # [17, 3] - (x, y, confidence)
+            # Get all keypoints and bounding boxes
+            keypoints = results[0].keypoints.data  # [num_people, 17, 3] - (x, y, confidence)
+            boxes = results[0].boxes.xyxy.cpu()  # [num_people, 4] - (x1, y1, x2, y2)
 
-            # Extract keypoints to landmarks_data
-            landmarks_data = []
-            for idx, kp in enumerate(keypoints):
-                x, y, conf = kp.tolist()
-                if conf >= detection_threshold:
-                    landmarks_data.append({
-                        'idx': idx,
-                        'x': x / w,  # Normalize to 0-1 range
-                        'y': y / h,  # Normalize to 0-1 range
-                        'confidence': conf
+            for i, (kps, box) in enumerate(zip(keypoints, boxes)):
+                # Extract keypoints to landmarks_data
+                landmarks_data = []
+                for idx, kp in enumerate(kps):
+                    x, y, conf = kp.tolist()
+                    if conf >= detection_threshold:
+                        landmarks_data.append({
+                            'idx': idx,
+                            'x': round(x / w, 4),  # Normalize to 0-1 range with 4 decimal precision
+                            'y': round(y / h, 4),  # Normalize to 0-1 range with 4 decimal precision
+                            'confidence': round(conf, 2)  # Reduce confidence to 2 decimal places
+                        })
+
+                if landmarks_data:  # Only add if we have valid landmarks
+                    # Add bounding box and landmarks for this person
+                    people_data.append({
+                        'bbox': box.tolist(),  # Store unnormalized for IoU calculation
+                        'landmarks': landmarks_data  # Store normalized for consistency
                     })
 
             # Create visualization if preview is enabled
             if show_preview:
                 processed_frame = results[0].plot()
 
-        return processed_frame, landmarks_data
+                # Add person IDs to the visualization if they're already assigned
+                for person in people_data:
+                    if 'person_id' in person:
+                        # Get center of bounding box
+                        x1, y1, x2, y2 = person['bbox']
+                        center_x = int((x1 + x2) / 2)
+                        center_y = int(y1)  # Top of the bbox
+
+                        # Draw ID text
+                        cv2.putText(
+                            processed_frame,
+                            f"ID: {person['person_id']}",
+                            (center_x, center_y - 10),
+                            cv2.FONT_HERSHEY_SIMPLEX,
+                            0.8,
+                            (0, 255, 255),
+                            2
+                        )
+
+        return processed_frame, people_data
 
     except RuntimeError as e:
         # Check if this is an NMS backend error
@@ -170,7 +422,9 @@ def run_pose_detection(
     model_size='n',
     device='auto',
     show_preview=True,
-    batch_size=1
+    batch_size=1,
+    frame_sampling=1,  # New parameter to control frame sampling rate
+    precision=3        # New parameter to control coordinate precision
 ):
     """YOLOv11 pose detection with CUDA acceleration, properly handling NMS issues"""
     start_time = time.time()
@@ -240,12 +494,13 @@ def run_pose_detection(
         window_name = "YOLOv11 Pose"
         cv2.namedWindow(window_name, cv2.WINDOW_NORMAL)
 
-    # Initialize variables for batch processing
-    all_landmarks = []
+    # Initialize variables for tracking
+    all_frame_data = []
     processed_frames = 0
-    frames_buffer = []
+    last_people_data = []
     last_fps_update = time.time()
     current_fps = 0
+    total_people_detected = 0
 
     # Main processing loop
     print("⏳ Processing frames...")
@@ -256,18 +511,23 @@ def run_pose_detection(
 
         try:
             # Process the frame
-            processed_frame, landmarks_data = process_frame(
+            processed_frame, people_data = process_frame(
                 frame, model, detection_threshold, show_preview
             )
 
-            # Store landmark data with timestamp
-            if landmarks_data:
+            # Assign stable person IDs
+            if people_data:
+                people_data = assign_person_ids(people_data, last_people_data)
+                last_people_data = people_data.copy()
+
+                # Store frame data with people
                 frame_data = {
                     'frame': processed_frames,
                     'timestamp': processed_frames / fps if fps > 0 else time.time() - start_time,
-                    'landmarks': landmarks_data
+                    'people': people_data
                 }
-                all_landmarks.append(frame_data)
+                all_frame_data.append(frame_data)
+                total_people_detected += len(people_data)
 
         except RuntimeError as e:
             if str(e) == "CUDA NMS Error":
@@ -295,7 +555,7 @@ def run_pose_detection(
             # Show CUDA status
             cv2.putText(
                 processed_frame,
-                f"Device: {model.device} (Full GPU processing)",
+                f"Device: {model.device} | People: {len(people_data) if people_data else 0}",
                 (10, 60), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2
             )
 
@@ -320,50 +580,62 @@ def run_pose_detection(
 
     print(f"⏱️ Processed {processed_frames} frames in {elapsed_time:.2f}s ({effective_fps:.2f} fps)")
 
-    if all_landmarks:
-        print(f"🧮 Detected poses in {len(all_landmarks)} frames ({(len(all_landmarks)/max(1, processed_frames))*100:.1f}%)")
+    if all_frame_data:
+        unique_people = set()
+        for frame in all_frame_data:
+            for person in frame['people']:
+                unique_people.add(person['person_id'])
+
+        print(f"🧮 Detected {len(all_frame_data)} frames with poses ({len(all_frame_data)/max(1, processed_frames)*100:.1f}%)")
+        print(f"👥 Detected {len(unique_people)} unique people with {total_people_detected} total detections")
     else:
         print(f"⚠️ No poses detected. Try adjusting detection threshold or check the video content.")
 
     # Save results if output file is specified
-    if output_file and all_landmarks:
+    if output_file and all_frame_data:
         output_dir = os.path.dirname(output_file)
         if output_dir:
             os.makedirs(output_dir, exist_ok=True)
 
         # Apply normalization if requested
-        if normalize and len(all_landmarks) > filter_window_size:
-            print(f"🔄 Normalizing data...")
-            landmarks_only = [frame_data['landmarks'] for frame_data in all_landmarks]
-            normalized_landmarks = normalize_landmarks(
-                landmarks_only,
+        if normalize and len(all_frame_data) > filter_window_size:
+            print(f"🔄 Normalizing data for each person...")
+            all_frame_data = normalize_landmarks_per_person(
+                all_frame_data,
                 window_size=filter_window_size,
                 poly_order=filter_poly_order
             )
 
-            # Put normalized landmarks back
-            for i, frame_data in enumerate(all_landmarks):
-                if i < len(normalized_landmarks):
-                    all_landmarks[i]['landmarks'] = normalized_landmarks[i]
+        # Compress data to reduce file size
+        print(f"🗜️ Compressing data (frame sampling: {frame_sampling}, precision: {precision})...")
+        compressed_frames = compress_pose_data(all_frame_data, frame_sampling, precision)
+        actual_frames_saved = len(compressed_frames)
 
-        # Create output in compatible format
+        # Calculate compression ratio
+        original_frame_count = len(all_frame_data)
+        compression_ratio = (original_frame_count - actual_frames_saved) / original_frame_count * 100
+        print(f"📊 Compression: {original_frame_count} frames reduced to {actual_frames_saved} ({compression_ratio:.1f}% reduction)")
+
+        # Create output in compatible format with compressed frames
         json_data = {
-            'source': source_name,
-            'frame_width': frame_width,
-            'frame_height': frame_height,
+            'src': source_name,  # Shortened key
+            'w': frame_width,    # Shortened key
+            'h': frame_height,   # Shortened key
             'fps': fps,
-            'total_frames': processed_frames,
-            'keypoint_names': KEYPOINT_NAMES,
-            'connections': [{'start': c[0], 'end': c[1]} for c in POSE_CONNECTIONS],
-            'frames': all_landmarks,
-            'metadata': {
+            'frames': processed_frames,
+            'keypoints': KEYPOINT_NAMES,  # More descriptive key
+            'connections': [{'s': c[0], 'e': c[1]} for c in POSE_CONNECTIONS],  # Shortened keys
+            'data': compressed_frames,  # Use compressed data
+            'meta': {  # Shortened key
                 'model': f"YOLOv11-{model_size}-pose",
                 'device': str(model.device),
                 'normalized': normalize,
-                'detection_threshold': detection_threshold,
-                'filter_window_size': filter_window_size if normalize else None,
-                'filter_poly_order': filter_poly_order if normalize else None,
-                'created_at': time.strftime('%Y-%m-%d %H:%M:%S')
+                'threshold': detection_threshold,
+                'filter_size': filter_window_size if normalize else None,
+                'filter_order': filter_poly_order if normalize else None,
+                'frame_sampling': frame_sampling,
+                'precision': precision,
+                'created': time.strftime('%Y-%m-%d %H:%M:%S')
             }
         }
 
@@ -371,7 +643,8 @@ def run_pose_detection(
         with open(output_file, 'w') as f:
             json.dump(json_data, f)
 
-        print(f"💾 Saved tracking data to {output_file}")
+        file_size_mb = os.path.getsize(output_file) / (1024 * 1024)
+        print(f"💾 Saved tracking data to {output_file} ({file_size_mb:.2f} MB)")
     elif output_file:
         print(f"⚠️ No pose data to save. Output file was not created.")
 
@@ -383,7 +656,7 @@ def run_pose_detection(
     # Restore original NMS function
     torchvision.ops.nms = original_nms
 
-    return all_landmarks
+    return all_frame_data
 
 def main():
     # Set up simple argument parser
@@ -411,8 +684,12 @@ def main():
                       help='Window size for smoothing filter (must be odd, larger = smoother)')
     parser.add_argument('--filter-order', type=int, default=4,
                       help='Polynomial order for smoothing filter (1-4)')
-    parser.add_argument('--batch-size', type=int, default=1,
+    parser.add_argument('--batch-size', type=int, default=4,
                       help='Batch size for processing (higher uses more VRAM but can be faster)')
+    parser.add_argument('--frame-sampling', type=int, default=2,
+                      help='Save only every Nth frame (1=all frames, 2=half, 4=quarter, etc.)')
+    parser.add_argument('--precision', type=int, default=3, choices=[2, 3, 4],
+                      help='Decimal precision for coordinates (2-4, lower=smaller file)')
 
     args = parser.parse_args()
 
@@ -430,6 +707,8 @@ def main():
     print(f"• Device: {args.device}")
     print(f"• Preview: {'Disabled' if args.no_preview else 'Enabled'}")
     print(f"• Normalization: {'Disabled' if args.no_normalize else 'Enabled'}")
+    print(f"• Frame sampling: Every {args.frame_sampling} frame(s)")
+    print(f"• Coordinate precision: {args.precision} decimal places")
     print("="*50 + "\n")
 
     # Run pose detection
@@ -444,7 +723,9 @@ def main():
             model_size=args.model,
             device=args.device,
             show_preview=not args.no_preview,
-            batch_size=args.batch_size
+            batch_size=args.batch_size,
+            frame_sampling=args.frame_sampling,
+            precision=args.precision
         )
     except KeyboardInterrupt:
         print("\n⏹️ Process interrupted by user")
diff --git a/pose_viewer.py b/pose_viewer.py
new file mode 100644
index 0000000..3c30feb
--- /dev/null
+++ b/pose_viewer.py
@@ -0,0 +1,436 @@
+import argparse
+import json
+import os
+import sys
+import time
+import urllib.request
+from pathlib import Path
+
+import cv2
+import numpy as np
+import pygame
+from pygame.locals import *
+
+# Define colors
+BLACK = (0, 0, 0)
+WHITE = (255, 255, 255)
+RED = (255, 0, 0)
+GREEN = (0, 255, 0)
+BLUE = (0, 0, 255)
+YELLOW = (255, 255, 0)
+CYAN = (0, 255, 255)
+MAGENTA = (255, 0, 255)
+
+# Define keypoint colors (custom palette)
+KEYPOINT_COLORS = [
+    (255, 0, 0),      # nose (red)
+    (255, 85, 0),     # left_eye (orange-red)
+    (255, 170, 0),    # right_eye (orange)
+    (255, 255, 0),    # left_ear (yellow)
+    (170, 255, 0),    # right_ear (yellow-green)
+    (85, 255, 0),     # left_shoulder (green-yellow)
+    (0, 255, 0),      # right_shoulder (green)
+    (0, 255, 85),     # left_elbow (green-cyan)
+    (0, 255, 170),    # right_elbow (cyan-green)
+    (0, 255, 255),    # left_wrist (cyan)
+    (0, 170, 255),    # right_wrist (cyan-blue)
+    (0, 85, 255),     # left_hip (blue-cyan)
+    (0, 0, 255),      # right_hip (blue)
+    (85, 0, 255),     # left_knee (blue-purple)
+    (170, 0, 255),    # right_knee (purple-blue)
+    (255, 0, 255),    # left_ankle (magenta)
+    (255, 0, 170)     # right_ankle (magenta-pink)
+]
+
+# Person ID colors
+PERSON_COLORS = [
+    (255, 0, 0),     # red
+    (0, 255, 0),     # green
+    (0, 0, 255),     # blue
+    (255, 255, 0),   # yellow
+    (255, 0, 255),   # magenta
+    (0, 255, 255),   # cyan
+    (255, 128, 0),   # orange
+    (128, 0, 255),   # purple
+    (0, 255, 128),   # mint
+    (255, 255, 255)  # white
+]
+
+def download_video(url, output_dir="downloaded_videos"):
+    """Download a video from a URL and return the local file path"""
+    os.makedirs(output_dir, exist_ok=True)
+    video_name = os.path.basename(url).split("?")[0]
+    if not video_name or "." not in video_name:
+        video_name = f"video_{int(time.time())}.mp4"
+
+    output_path = os.path.join(output_dir, video_name)
+    if os.path.exists(output_path):
+        print(f"✅ Video already downloaded: {output_path}")
+        return output_path
+
+    print(f"⬇️ Downloading video from {url} to {output_path}...")
+    urllib.request.urlretrieve(url, output_path)
+    print(f"✅ Video downloaded successfully to {output_path}")
+    return output_path
+
+def load_pose_data(json_file):
+    """Load pose data from a JSON file"""
+    print(f"📂 Loading pose data from {json_file}...")
+    with open(json_file, 'r') as f:
+        data = json.load(f)
+
+    # Extract metadata
+    width = data.get('w', 1280)
+    height = data.get('h', 720)
+    fps = data.get('fps', 30)
+    total_frames = data.get('frames', 0)
+
+    # Get frame sampling and precision from metadata if available
+    metadata = data.get('meta', {})
+    frame_sampling = metadata.get('frame_sampling', 1)
+    precision = metadata.get('precision', 3)
+
+    # Extract connections
+    connections = []
+    for conn in data.get('connections', []):
+        start = conn.get('s', 0)
+        end = conn.get('e', 0)
+        connections.append((start, end))
+
+    # Extract keypoint names
+    keypoint_names = data.get('keypoints', [])
+
+    # Extract frame data
+    frames = data.get('data', [])
+
+    print(f"✅ Loaded {len(frames)} frames of pose data")
+    print(f"📊 Video: {width}x{height}@{fps}fps, {total_frames} total frames")
+    print(f"🔍 Frame sampling: {frame_sampling}, Precision: {precision}")
+
+    return {
+        'width': width,
+        'height': height,
+        'fps': fps,
+        'total_frames': total_frames,
+        'frame_sampling': frame_sampling,
+        'precision': precision,
+        'connections': connections,
+        'keypoint_names': keypoint_names,
+        'frames': frames
+    }
+
+def create_pygame_window(width, height, title="Pose Viewer"):
+    """Create a PyGame window"""
+    pygame.init()
+    window = pygame.display.set_mode((width, height))
+    pygame.display.set_caption(title)
+    return window
+
+def draw_pose(frame, pose_data, frame_idx, original_width, original_height, prev_frame_idx=None):
+    """Draw pose data on a given frame"""
+    # Clone the frame to avoid modifying the original
+    pose_frame = frame.copy()
+
+    # Find the closest pose frame to the current video frame (should be first and only in temp_pose_data)
+    if pose_data['frames']:
+        closest_frame = pose_data['frames'][0]
+
+        connections = pose_data['connections']
+
+        # Draw each person
+        for person_idx, person in enumerate(closest_frame['p']):
+            person_id = person['id']
+            person_color = PERSON_COLORS[person_id % len(PERSON_COLORS)]
+
+            # Get keypoints
+            keypoints = person['k']
+
+            # Create a dictionary to store keypoints by index
+            kp_dict = {}
+            for kp in keypoints:
+                kp_dict[kp[0]] = (
+                    int(kp[1] * original_width),
+                    int(kp[2] * original_height),
+                    kp[3]
+                )
+
+            # Draw connections
+            for conn in connections:
+                if conn[0] in kp_dict and conn[1] in kp_dict:
+                    start_point = kp_dict[conn[0]][:2]
+                    end_point = kp_dict[conn[1]][:2]
+
+                    # Use average confidence to determine line thickness
+                    avg_conf = (kp_dict[conn[0]][2] + kp_dict[conn[1]][2]) / 2
+                    thickness = int(avg_conf * 3) + 1
+
+                    cv2.line(pose_frame, start_point, end_point, person_color, thickness)
+
+            # Draw keypoints
+            for kp_idx, (x, y, conf) in kp_dict.items():
+                # Circle size based on confidence
+                radius = int(conf * 5) + 2
+                cv2.circle(pose_frame, (x, y), radius, KEYPOINT_COLORS[kp_idx % len(KEYPOINT_COLORS)], -1)
+
+            # Draw person ID
+            bbox = person['b']
+            x, y = int(bbox[0]), int(bbox[1])
+            cv2.putText(
+                pose_frame,
+                f"ID: {person_id}",
+                (x, y - 10),
+                cv2.FONT_HERSHEY_SIMPLEX,
+                0.7,
+                person_color,
+                2
+            )
+
+    return pose_frame
+
+def draw_ui_controls(surface, width, height, playing, current_frame, total_frames):
+    """Draw UI controls on the PyGame surface"""
+    # Background for controls
+    control_height = 50
+    control_surface = pygame.Surface((width, control_height))
+    control_surface.fill(BLACK)
+
+    # Draw play/pause button
+    button_width = 80
+    button_height = 30
+    button_x = 20
+    button_y = (control_height - button_height) // 2
+
+    pygame.draw.rect(control_surface, BLUE, (button_x, button_y, button_width, button_height))
+
+    font = pygame.font.SysFont(None, 24)
+    text = font.render("Pause" if playing else "Play", True, WHITE)
+    text_rect = text.get_rect(center=(button_x + button_width//2, button_y + button_height//2))
+    control_surface.blit(text, text_rect)
+
+    # Draw stop button
+    stop_button_x = button_x + button_width + 20
+    pygame.draw.rect(control_surface, RED, (stop_button_x, button_y, button_width, button_height))
+
+    stop_text = font.render("Stop", True, WHITE)
+    stop_text_rect = stop_text.get_rect(center=(stop_button_x + button_width//2, button_y + button_height//2))
+    control_surface.blit(stop_text, stop_text_rect)
+
+    # Draw seek bar
+    seekbar_x = stop_button_x + button_width + 40
+    seekbar_y = button_y + button_height // 2
+    seekbar_width = width - seekbar_x - 40
+    seekbar_height = 10
+
+    # Background bar
+    pygame.draw.rect(control_surface, (100, 100, 100),
+                    (seekbar_x, seekbar_y - seekbar_height//2, seekbar_width, seekbar_height))
+
+    # Progress bar
+    progress = current_frame / total_frames if total_frames > 0 else 0
+    progress_width = int(seekbar_width * progress)
+    pygame.draw.rect(control_surface, GREEN,
+                    (seekbar_x, seekbar_y - seekbar_height//2, progress_width, seekbar_height))
+
+    # Display current time / total time
+    time_text = font.render(f"Frame: {current_frame} / {total_frames}", True, WHITE)
+    time_rect = time_text.get_rect(center=(seekbar_x + seekbar_width//2, seekbar_y - 20))
+    control_surface.blit(time_text, time_rect)
+
+    # Blit the control surface to the main surface
+    surface.blit(control_surface, (0, height - control_height))
+
+    # Return button regions for click handling
+    play_button_rect = pygame.Rect(button_x, height - control_height + button_y, button_width, button_height)
+    stop_button_rect = pygame.Rect(stop_button_x, height - control_height + button_y, button_width, button_height)
+    seekbar_rect = pygame.Rect(seekbar_x, height - control_height + seekbar_y - seekbar_height//2,
+                              seekbar_width, seekbar_height)
+
+    return play_button_rect, stop_button_rect, seekbar_rect
+
+def run_viewer(video_path, json_path):
+    """Main function to run the pose viewer"""
+    # Load pose data
+    pose_data = load_pose_data(json_path)
+
+    # Open video
+    cap = cv2.VideoCapture(video_path)
+    if not cap.isOpened():
+        print(f"❌ Error: Could not open video {video_path}")
+        return
+
+    # Get video properties
+    video_width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+    video_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+    fps = cap.get(cv2.CAP_PROP_FPS)
+    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
+
+    # Create window - now only showing single visualization
+    window_width = video_width
+    window_height = video_height + 50  # Additional space for controls
+    window = create_pygame_window(window_width, window_height, f"Pose Viewer - {os.path.basename(video_path)}")
+
+    # Setup clock
+    clock = pygame.time.Clock()
+
+    # State variables
+    playing = False
+    current_frame = 0
+    prev_frame_idx = None
+    last_valid_pose_frame = None
+
+    # Create a frame lookup for fast access - maps video frame number to pose data frame
+    frame_lookup = {}
+    max_pose_frame = 0
+    for pose_frame in pose_data['frames']:
+        frame_num = pose_frame['f']
+        frame_lookup[frame_num] = pose_frame
+        max_pose_frame = max(max_pose_frame, frame_num)
+
+    # Initial render
+    ret, frame = cap.read()
+    if not ret:
+        print("❌ Error: Could not read the first frame")
+        return
+
+    # Main loop
+    while True:
+        # Handle events
+        for event in pygame.event.get():
+            if event.type == QUIT:
+                pygame.quit()
+                sys.exit()
+            elif event.type == KEYDOWN:
+                if event.key == K_ESCAPE:
+                    pygame.quit()
+                    sys.exit()
+                elif event.key == K_SPACE:
+                    playing = not playing
+            elif event.type == MOUSEBUTTONDOWN:
+                # Check if any buttons were clicked
+                mouse_pos = pygame.mouse.get_pos()
+                play_button_rect, stop_button_rect, seekbar_rect = draw_ui_controls(
+                    window, window_width, window_height, playing, current_frame, total_frames
+                )
+
+                if play_button_rect.collidepoint(mouse_pos):
+                    playing = not playing
+                elif stop_button_rect.collidepoint(mouse_pos):
+                    playing = False
+                    current_frame = 0
+                    cap.set(cv2.CAP_PROP_POS_FRAMES, current_frame)
+                    ret, frame = cap.read()
+                    prev_frame_idx = None
+                    last_valid_pose_frame = None
+                elif seekbar_rect.collidepoint(mouse_pos):
+                    # Calculate position ratio
+                    x_offset = mouse_pos[0] - seekbar_rect.x
+                    ratio = x_offset / seekbar_rect.width
+
+                    # Set frame position
+                    current_frame = int(ratio * total_frames)
+                    cap.set(cv2.CAP_PROP_POS_FRAMES, current_frame)
+                    ret, frame = cap.read()
+                    prev_frame_idx = None  # Reset previous frame index after seeking
+
+        # Handle playback
+        if playing:
+            ret, frame = cap.read()
+            if not ret:
+                # End of video, loop back to start
+                playing = False
+                current_frame = 0
+                cap.set(cv2.CAP_PROP_POS_FRAMES, current_frame)
+                ret, frame = cap.read()
+                prev_frame_idx = None
+                last_valid_pose_frame = None
+                if not ret:
+                    break
+            current_frame += 1
+
+        # Find the appropriate pose frame for the current video frame
+        frame_sampling = pose_data['frame_sampling']
+
+        # Try to find the exact frame in lookup
+        pose_frame = frame_lookup.get(current_frame)
+
+        # If not found, find the closest previous frame based on sampling
+        if not pose_frame:
+            # Calculate what the nearest pose frame should be
+            # This searches for the most recent pose frame
+            nearest_frame = current_frame
+            while nearest_frame > 0 and nearest_frame not in frame_lookup:
+                nearest_frame -= 1
+
+            if nearest_frame in frame_lookup:
+                pose_frame = frame_lookup[nearest_frame]
+
+        # Update the last valid pose frame if we found one
+        if pose_frame:
+            last_valid_pose_frame = pose_frame
+
+        # Draw pose on frame - use the most recent valid pose frame
+        if last_valid_pose_frame:
+            # Create a special frame dict with only the current pose for draw_pose
+            temp_pose_data = pose_data.copy()
+            temp_pose_data['frames'] = [last_valid_pose_frame]
+
+            pose_frame = draw_pose(frame, temp_pose_data, current_frame, video_width, video_height)
+        else:
+            # If no pose data found yet, just show the original frame
+            pose_frame = frame.copy()
+
+        prev_frame_idx = current_frame
+
+        # Convert frame from BGR to RGB for PyGame
+        rgb_pose_frame = cv2.cvtColor(pose_frame, cv2.COLOR_BGR2RGB)
+        pygame_pose_frame = pygame.surfarray.make_surface(rgb_pose_frame.swapaxes(0, 1))
+
+        # Draw frame
+        window.blit(pygame_pose_frame, (0, 0))
+
+        # Draw UI controls
+        play_button_rect, stop_button_rect, seekbar_rect = draw_ui_controls(
+            window, window_width, window_height, playing, current_frame, total_frames
+        )
+
+        # Draw metadata
+        font = pygame.font.SysFont(None, 20)
+        metadata_text = f"Frame Sampling: {pose_data['frame_sampling']}, Precision: {pose_data['precision']}"
+        metadata_surface = font.render(metadata_text, True, WHITE)
+        window.blit(metadata_surface, (10, 10))
+
+        # Update display
+        pygame.display.flip()
+
+        # Cap framerate
+        clock.tick(fps)
+
+    # Clean up
+    cap.release()
+    pygame.quit()
+
+def main():
+    parser = argparse.ArgumentParser(description='Pose Viewer for JSON pose data with video')
+    parser.add_argument('--video', '-v', required=True, help='Video file path or URL')
+    parser.add_argument('--json', '-j', required=True, help='JSON pose data file path')
+
+    args = parser.parse_args()
+
+    # Handle URL input for video
+    video_path = args.video
+    if video_path.startswith('http://') or video_path.startswith('https://'):
+        video_path = download_video(video_path)
+
+    if not os.path.exists(video_path):
+        print(f"❌ Error: Video file not found: {video_path}")
+        return
+
+    if not os.path.exists(args.json):
+        print(f"❌ Error: JSON file not found: {args.json}")
+        return
+
+    # Run the viewer
+    run_viewer(video_path, args.json)
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
index 736c024..0f15e17 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -6,4 +6,5 @@ flask-cors>=3.0.10
 numpy>=1.19.0
 scipy>=1.7.0
 pillow>=9.0.0
-mediapipe>=0.8.9
\ No newline at end of file
+mediapipe>=0.8.9
+pygame>=2.0.0
\ No newline at end of file