Update app.py

app.py

@@ -1,168 +1,55 @@
-import random 
-import gradio as gr 
-import numpy as np 
-from elevenlabs import voices, generate, set_api_key, UnauthenticatedRateLimitError
+import gradio as gr
+import subprocess
+import shutil
+import os
+from PIL import Image
+import numpy as np
+
+# Ensure output directory exists
+output_dir = "output"
+if not os.path.exists(output_dir):
+    os.makedirs(output_dir)
 
-def pad_buffer(audio):
-    # Pad buffer to multiple of 2 bytes
-    buffer_size = len(audio)
-    element_size = np.dtype(np.int16).itemsize
-    if buffer_size % element_size != 0:
-        audio = audio + b'\0' * (element_size - (buffer_size % element_size))
-    return audio 
-
-def generate_voice(text, voice_name):
-    try:
-        audio = generate(
-            text[:250], # Limit to 250 characters
-            voice=voice_name, 
-            model="eleven_multilingual_v2"
-        )
-        return (44100, np.frombuffer(pad_buffer(audio), dtype=np.int16))
-    except UnauthenticatedRateLimitError as e:
-        raise gr.Error("Thanks for trying out ElevenLabs TTS! You've reached the free tier limit. Please provide an API key to continue.") 
-    except Exception as e:
-        raise gr.Error(e)
-    
 # Login function for authentication
 def custom_auth(username, password):
     return password == "aitutor"
 
-
-badges = """
-"""
-menu = """
-<style>
-    /* Existing styles */
-    .gradio-container-3-40-1 .prose a {
-        color: #ffffff!important;
-        text-decoration: auto!important;
-    }
-    .menu-bar {
-        display: flex;
-        justify-content: space-between;
-        align-items: center;
-        background-color: #333;
-        padding: 10px;
-        color: white;
-        font-family: 'sans-serif';
-    }
-    .menu-bar a, .menu-bar a:visited {
-        color: white;
-        text-decoration: none;
-    }
-    .menu-icon {
-        font-size: 24px;
-        background-color: #ffffff;
-        border-radius: 50%;
-        padding: 5px;
-    }
-    .menu-items {
-        display: flex;
-        gap: 15px;
-        color: white;
-    }
-    .menu-item {
-        padding: 8px 16px;
-        background-color: #555;
-        border-radius: 4px;
-        transition: background-color 0.3s;
-        font-weight: bold;
-        font-size: 12px;
-    }
-    .menu-item:hover {
-        background-color: #777;
-    }
-
-    /* Responsive styles for mobile */
-    @media (max-width: 768px) {
-        .menu-item {
-            font-size: 12px;  /* Shrink text size */
-        }
-        .menu-icon {
-            font-size: 18px;  /* Shrink icon size */
-        }
-    }
-</style>
-
-<div class="menu-bar">
-    <a href="#" class="menu-icon">🎵</a>
-    <div class="menu-items">
-        <span class="menu-item"><a href="#">Dashboard</a></span>
-        <span class="menu-item"><a href="#">Premium</a></span>
-        <span class="menu-item"><a href="#">Account</a></span>
-        <span class="menu-item"><a href="#">Voices</a></span>
-    </div>
-</div>
-
-"""
-
-description = """
-<style>
-    .notification {
-        text-align: left;  /* Left-align the text */
-        background-color: #ffffff;  /* White Background */
-        color: #000;  /* Black text */
-        padding: 20px;  /* Padding */
-        margin: 20px;  /* Margin */
-        margin-bottom: 40px;  /* Additional space below the notification */
-        border: 2px solid #000000;  /* Black border */
-        border-radius: 20px;  /* Rounded corners */
-        font-size: 18px;  /* Font size */
-        font-family: 'Arial', sans-serif;  /* Font family */
-        box-shadow: 0 4px 8px 0 rgba(0,0,0,0.1);  /* Subtle box shadow for depth */
-    }
-</style>
-
-<div class="notification">
-    Welcome to Text to Voice, single foundational model supporting 28 languages including: English, Chinese, Spanish, Hindi, Portuguese, French, German, Japanese, Arabic, Korean, Indonesian, Italian, Dutch, Turkish, Polish, Swedish, Filipino, Malay, Romanian, Ukrainian, Greek, Czech, Danish, Finnish, Bulgarian, Croatian, Slovak, and Tamil.
-</div>
-
-"""
-
-
-
-with gr.Blocks() as block:
-    gr.Markdown('')
-    gr.Markdown(menu)
-    gr.Markdown(badges)
-    gr.Markdown(description)
-
-    input_text = gr.Textbox(
-        label="Input Text (250 characters max)",
-        lines=2,
-        value="Hello! 你好! Hola! नमस्ते! Bonjour! こんにちは! مرحبا! 안녕하세요! Ciao! Cześć! Привіт! Γειά σας! Здравей! வணக்கம்!",
-        elem_id="input_text"
-    )
-
-    all_voices = voices()
-    input_voice = gr.Dropdown(
-        [voice.name for voice in all_voices],
-        value="Bella",
-        label="Voice",
-        elem_id="input_voice"
-    )
-
-    run_button = gr.Button(
-        text="Generate Voice",
-        type="button"
-    )
-
-    out_audio = gr.Audio(
-        label="Generated Voice",
-        type="numpy",
-        elem_id="out_audio",
-        format="mp3"
-    )
-
-    inputs = [input_text, input_voice]
-    outputs = [out_audio]
-
-    run_button.click(
-        fn=generate_voice,
-        inputs=inputs,
-        outputs=outputs,
-        queue=True
-    )
-
-block.queue(concurrency_count=5).launch(debug=True, auth=custom_auth)
+def run_scripts(target, source, use_face_enhancer):
+    if target is None or (not use_face_enhancer and source is None):
+        return None
+
+    target_extension = os.path.splitext(target.name)[-1]
+    output_path1 = os.path.join(output_dir, "output1" + target_extension)
+    output_path2 = os.path.join(output_dir, "output2" + target_extension)
+
+    if not use_face_enhancer:
+        cmd1 = ["python3", "run.py", "-s", source.name, "-t", target.name, "-o", output_path1, "--frame-processor", "face_swapper"]
+        subprocess.run(cmd1)
+
+    cmd2 = ["python3", "run.py", "-t", target.name if use_face_enhancer else output_path1, "-o", output_path2, "--frame-processor", "face_enhancer"]
+    subprocess.run(cmd2)
+
+    if not use_face_enhancer:
+        os.remove(source.name)
+    os.remove(target.name)
+
+    # Open the image with PIL and convert to NumPy array
+    with Image.open(output_path2) as img:
+        img_array = np.array(img)
+
+    return img_array
+
+iface = gr.Interface(
+    fn=run_scripts,
+    inputs=[
+        "file",
+        "file",
+        gr.inputs.Checkbox(default=False, label="Use only Face Enhancer")
+    ],
+    outputs="image",
+    title="Pixio Swap",
+    description="Upload a target image/video and a source image to swap faces.",
+    live=True
+)
+
+iface.launch(debug=True, auth=custom_auth)