Create app.py

app.py

@@ -0,0 +1,58 @@
+import torch, cv2, numpy as np
+from diffusers import StableDiffusionPipeline
+import moviepy.editor as mp
+from gtts import gTTS
+import gradio as gr
+
+# ---------- Image ----------
+def generate_image(prompt, ratio="16:9"):
+    model = StableDiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-2-1").to("cuda")
+    image = model(prompt).images[0]
+    return image
+
+# ---------- Video ----------
+def generate_video(prompt, duration=12, ratio="16:9"):
+    model = StableDiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-2-1").to("cuda")
+    frames = []
+    for _ in range(duration):
+        img = model(prompt).images[0]
+        frame = cv2.cvtColor(np.array(img), cv2.COLOR_RGB2BGR)
+        frames.append(frame)
+
+    h, w, _ = frames[0].shape
+    out = cv2.VideoWriter("output_video.mp4", cv2.VideoWriter_fourcc(*"mp4v"), 1, (w, h))
+    for f in frames: out.write(f)
+    out.release()
+    return "output_video.mp4"
+
+# ---------- Voice ----------
+def generate_voice(text, lang="bn"):
+    tts = gTTS(text=text, lang=lang)
+    filename = f"voice_{lang}.mp3"
+    tts.save(filename)
+    return filename
+
+# ---------- Gradio UI ----------
+with gr.Blocks() as demo:
+    gr.Markdown("# 🎬 AI App → Image | Video | Voice")
+
+    with gr.Tab("🖼️ Image"):
+        prompt_img = gr.Textbox(label="Prompt", value="3D futuristic city in 4K")
+        btn_img = gr.Button("Generate Image")
+        output_img = gr.Image()
+        btn_img.click(generate_image, inputs=prompt_img, outputs=output_img)
+
+    with gr.Tab("🎥 Video"):
+        prompt_vid = gr.Textbox(label="Prompt", value="Flying car 3D animation 4K")
+        btn_vid = gr.Button("Generate Video (10–15s)")
+        output_vid = gr.Video()
+        btn_vid.click(generate_video, inputs=prompt_vid, outputs=output_vid)
+
+    with gr.Tab("🎙️ Voice"):
+        text_voice = gr.Textbox(label="Text", value="এটি একটি টেস্ট ভয়েস")
+        lang_voice = gr.Dropdown(choices=["bn","en","hi","ar"], label="Language", value="bn")
+        btn_voice = gr.Button("Generate Voice")
+        output_voice = gr.Audio()
+        btn_voice.click(generate_voice, inputs=[text_voice, lang_voice], outputs=output_voice)
+
+demo.launch()