Update app.py

app.py

@@ -1,47 +1,63 @@
 import gradio as gr
 from transformers import pipeline
 from gtts import gTTS
-from moviepy.editor import ImageSequenceClip, CompositeVideoClip, ImageClip, AudioFileClip, VideoFileClip, TextClip, concatenate_videoclips
-from PIL import Image, ImageDraw, ImageFont
+from moviepy.editor import CompositeVideoClip, ImageClip, AudioFileClip, VideoFileClip, concatenate_videoclips
+from diffusers import StableDiffusionPipeline
+import torch
+from PIL import Image
 import numpy as np
 from scipy.io.wavfile import write
 
 # Load and Initialize Models
 script_generator = pipeline("text-generation", model="gpt2", truncation=True, max_length=100)
+image_generator = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", torch_dtype=torch.float16).to("cuda")
+music_generator = pipeline("text-to-music", model="HuggingFace/MusicGen")  # Replace with actual model if available
 
+# Generate Comedy Script
 def generate_comedy_script(prompt):
     script = script_generator(prompt)[0]['generated_text']
     return script
 
+# Convert Text to Speech
 def text_to_speech(script):
     tts = gTTS(text=script, lang='en')
     audio_file = 'output.mp3'
     tts.save(audio_file)
     return audio_file
 
+# Create Images Using Stable Diffusion
 def create_images_from_script(script):
     lines = script.split('. ')
     image_paths = []
     for i, line in enumerate(lines):
-        img = Image.new('RGB', (800, 400), color=(0, 0, 0))
-        d = ImageDraw.Draw(img)
-        fnt = ImageFont.load_default()
-        d.text((10, 180), line, font=fnt, fill=(255, 255, 255))
-        image_path = f'/tmp/image_{i}.png'
-        img.save(image_path)
-        image_paths.append(image_path)
+        img = image_generator(line).images[0]
+        img_path = f'/tmp/image_{i}.png'
+        img.save(img_path)
+        image_paths.append(img_path)
     return image_paths
 
-def generate_animation_with_images(script):
+# Generate Fun Music Track
+def generate_fun_music(prompt, output_music_file="fun_music.wav"):
+    # Generate music based on the prompt
+    music = music_generator(prompt)[0]['generated_music']
+    # Save the generated music to a file
+    with open(output_music_file, 'wb') as f:
+        f.write(music)
+    return output_music_file
+
+# Create Video from Generated Images
+def generate_text_video(script):
     image_paths = create_images_from_script(script)
     clips = []
-    for i, img_path in enumerate(image_paths):
+    for img_path in image_paths:
         image_clip = ImageClip(img_path).set_duration(3).set_position(('center', 'center'))
         clips.append(image_clip)
+    
     final_video = concatenate_videoclips(clips, method="compose")
     final_video.write_videofile("/tmp/final_video.mp4", fps=24)
     return "/tmp/final_video.mp4"
 
+# Combine Audio and Video
 def combine_audio_video(video_path, audio_path):
     video = VideoFileClip(video_path)
     audio = AudioFileClip(audio_path)
@@ -49,78 +65,24 @@ def combine_audio_video(video_path, audio_path):
     final_video.write_videofile("/tmp/final_comedy_video.mp4", fps=24)
     return "/tmp/final_comedy_video.mp4"
 
+# Main Function to Generate Comedy Animation
 def generate_comedy_and_animation(prompt):
     script = generate_comedy_script(prompt)
     audio_file = text_to_speech(script)
-    video_file = generate_animation_with_images(script)
-    final_video = combine_audio_video(video_file, audio_file)
+    video_file = generate_text_video(script)
+    fun_music = generate_fun_music(prompt)
+    final_video = combine_audio_video(video_file, fun_music)
     return script, audio_file, final_video
 
-def generate_sine_wave(frequency, duration, sample_rate=44100, amplitude=0.5):
-    t = np.linspace(0, duration, int(sample_rate * duration), False)
-    wave = amplitude * np.sin(2 * np.pi * frequency * t)
-    return wave
-
-def generate_kids_music(theme, output_music_file="kids_music.wav"):
-    sample_rate = 44100
-    duration = 2
-    c4_wave = generate_sine_wave(261.63, duration)
-    d4_wave = generate_sine_wave(293.66, duration)
-    e4_wave = generate_sine_wave(329.63, duration)
-    wave = np.concatenate([c4_wave, d4_wave, e4_wave])
-    audio_wave = np.int16(wave * 32767)
-    write(output_music_file, sample_rate, audio_wave)
-    return output_music_file
-
+# Generate Kids Music Animation
 def generate_kids_animation_with_music(theme, output_video_file="kids_animation.mp4"):
-    music_file = generate_kids_music(theme)
+    music_file = generate_fun_music(theme, output_music_file="kids_music.wav")
     clips = []
     for i in range(5):
         img = Image.new('RGB', (800, 400), color=(0, 0, 255))
         d = ImageDraw.Draw(img)
         fnt = ImageFont.load_default()
-        d.text((10, 180), f"Kids Music: {theme}", font=fnt, fill=(255, 255, 0))
-        frame_path = f'/tmp/kids_temp_{i}.png'
-        img.save(frame_path)
-        clips.append(ImageClip(frame_path).set_duration(1).set_position(('center', 'center')))
-    final_video = CompositeVideoClip(clips, size=(800, 400))
-    final_video = final_video.set_audio(AudioFileClip(music_file))
-    final_video.write_videofile(output_video_file, fps=24)
-    return output_video_file, music_file
-
-def generate_kids_content(theme):
-    video_file, music_file = generate_kids_animation_with_music(theme)
-    return music_file, video_file
-
-with gr.Blocks() as app:
-    gr.Markdown("## AI Comedy and Kids Content Generator")
-
-    with gr.Tab("Generate Comedy Animation"):
-        prompt_input = gr.Textbox(label="Comedy Prompt")
-        generate_btn = gr.Button("Generate Comedy Script and Animation")
-        comedy_script = gr.Textbox(label="Generated Script")
-        comedy_audio = gr.Audio(label="Generated Audio")
-        comedy_video = gr.Video(label="Generated Animation")
-
-        generate_btn.click(
-            generate_comedy_and_animation,
-            inputs=prompt_input,
-            outputs=[comedy_script, comedy_audio, comedy_video]
-        )
-
-    with gr.Tab("Generate Kids Music Animation"):
-        theme_input = gr.Textbox(label="Kids Music Theme")
-        generate_music_btn = gr.Button("Generate Kids Music and Animation")
-        kids_music_audio = gr.Audio(label="Generated Music")
-        kids_music_video = gr.Video(label="Generated Kids Animation")
-
-        generate_music_btn.click(
-            generate_kids_content,
-            inputs=theme_input,
-            outputs=[kids_music_audio, kids_music_video]
-        )
-
-app.launch()
+        d.text((10, 180