app.py

import glob
import math
import os
import tempfile
import time
from pathlib import Path
from typing import List, Optional, Tuple, Union
import subprocess
from dataclasses import dataclass

import gradio as gr
import matplotlib.pyplot as plt
import numpy as np
import torch
from loguru import logger
from PIL import Image as PILImage
from torch import Tensor
from torchaudio.backend.common import AudioMetaData

from df import config
from df.enhance import enhance, init_df, load_audio, save_audio
from df.io import resample

# ============================================================================
# Configuration and Setup
# ============================================================================

@dataclass
class AppConfig:
    """Application configuration"""
    device: torch.device
    sample_rate: int = 48000
    max_duration_seconds: int = 3600
    cleanup_hours: int = 2
    temp_dir: str = "/tmp"
    model_path: str = "./DeepFilterNet2"
    fade_duration: float = 0.15
    

class AudioProcessor:
    """Handles audio processing operations"""
    
    def __init__(self, model, df, config: AppConfig):
        self.model = model
        self.df = df
        self.config = config
        
    def mix_at_snr(self, clean: Tensor, noise: Tensor, snr: float, eps: float = 1e-10) -> Tuple[Tensor, Tensor, Tensor]:
        """Mix clean and noise signal at a given SNR with improved error handling."""
        clean = torch.as_tensor(clean).mean(0, keepdim=True)
        noise = torch.as_tensor(noise).mean(0, keepdim=True)
        
        if noise.shape[1] < clean.shape[1]:
            repeats = int(math.ceil(clean.shape[1] / noise.shape[1]))
            noise = noise.repeat((1, repeats))
        
        max_start = int(noise.shape[1] - clean.shape[1])
        start = torch.randint(0, max_start, ()).item() if max_start > 0 else 0
        noise = noise[:, start : start + clean.shape[1]]
        
        E_speech = torch.mean(clean.pow(2)) + eps
        E_noise = torch.mean(noise.pow(2)) + eps
        K = torch.sqrt((E_noise / E_speech) * 10 ** (snr / 10) + eps)
        noise = noise / K
        mixture = clean + noise
        
        assert torch.isfinite(mixture).all(), "Non-finite values detected in mixture"
        max_m = mixture.abs().max()
        if max_m > 1:
            logger.warning(f"Clipping detected during mixing. Reducing gain by {1/max_m:.3f}")
            clean, noise, mixture = clean / max_m, noise / max_m, mixture / max_m
            
        return clean, noise, mixture
    
    def enhance_audio(self, audio: Tensor) -> Tensor:
        """Enhance audio using the DeepFilterNet model."""
        logger.info(f"Enhancing audio with shape {audio.shape}")
        with torch.no_grad():
            enhanced = enhance(self.model, self.df, audio)
        
        sr = self.config.sample_rate
        fade_samples = int(sr * self.config.fade_duration)
        lim = torch.linspace(0.0, 1.0, fade_samples).unsqueeze(0)
        lim = torch.cat((lim, torch.ones(1, enhanced.shape[1] - lim.shape[1])), dim=1)
        enhanced = enhanced * lim
        
        return enhanced


class AudioLoader:
    """Handles audio loading from various sources"""
    
    @staticmethod
    def ensure_wav(filepath: str) -> str:
        """Convert audio files to WAV using ffmpeg if needed."""
        if not filepath:
            return filepath
            
        file_ext = Path(filepath).suffix.lower()
        if file_ext in ['.mp3', '.m4a', '.ogg', '.flac', '.aac']:
            wav_path = str(Path(filepath).with_suffix('.wav'))
            try:
                subprocess.run(
                    ["ffmpeg", "-y", "-i", filepath, "-acodec", "pcm_s16le", wav_path],
                    check=True,
                    capture_output=True
                )
                logger.info(f"Converted {file_ext} to WAV: {wav_path}")
                return wav_path
            except subprocess.CalledProcessError as e:
                logger.error(f"FFmpeg conversion failed: {e.stderr}")
                raise
        return filepath
    
    @staticmethod
    def load_audio_gradio(
        audio_or_file: Union[None, str, Tuple[int, np.ndarray]], 
        sr: int
    ) -> Optional[Tuple[Tensor, AudioMetaData]]:
        """Load audio from Gradio input."""
        if audio_or_file is None:
            return None
            
        if isinstance(audio_or_file, str):
            if audio_or_file.lower() == "none":
                return None
            audio_or_file = AudioLoader.ensure_wav(audio_or_file)
            audio, meta = load_audio(audio_or_file, sr)
        else:
            meta = AudioMetaData(-1, -1, -1, -1, "")
            assert isinstance(audio_or_file, (tuple, list))
            meta.sample_rate, audio_np = audio_or_file
            
            audio_np = audio_np.reshape(audio_np.shape[0], -1).T
            
            if audio_np.dtype == np.int16:
                audio_np = (audio_np / (1 << 15)).astype(np.float32)
            elif audio_np.dtype == np.int32:
                audio_np = (audio_np / (1 << 31)).astype(np.float32)
            
            audio = resample(torch.from_numpy(audio_np), meta.sample_rate, sr)
            
        return audio, meta


class SpectrogramVisualizer:
    """Handles spectrogram visualization"""
    
    def __init__(self, figsize: Tuple[float, float] = (15.2, 4)):
        self.figsize = figsize
        plt.style.use('dark_background')
        self.fig_noisy, self.ax_noisy = plt.subplots(figsize=figsize)
        self.fig_noisy.set_tight_layout(True)
        self.fig_enh, self.ax_enh = plt.subplots(figsize=figsize)
        self.fig_enh.set_tight_layout(True)
    
    def specshow(
        self,
        spec: Union[Tensor, np.ndarray],
        ax: Optional[plt.Axes] = None,
        title: Optional[str] = None,
        xlabel: Optional[str] = None,
        ylabel: Optional[str] = None,
        sr: int = 48000,
        n_fft: Optional[int] = None,
        hop: Optional[int] = None,
        vmin: float = -100,
        vmax: float = 0,
        cmap: str = "plasma",
    ):
        """Plot a spectrogram of shape [F, T]"""
        spec_np = spec.cpu().numpy() if isinstance(spec, torch.Tensor) else spec
        
        if n_fft is None:
            n_fft = spec.shape[0] * 2 if spec.shape[0] % 2 == 0 else (spec.shape[0] - 1) * 2
        hop = hop or n_fft // 4
        
        t = np.arange(0, spec_np.shape[-1]) * hop / sr
        f = np.arange(0, spec_np.shape[0]) * sr // 2 / (n_fft // 2) / 1000
        
        im = ax.pcolormesh(
            t, f, spec_np, 
            rasterized=True, 
            shading="auto", 
            vmin=vmin, 
            vmax=vmax, 
            cmap=cmap
        )
        
        if title:
            ax.set_title(title, fontsize=14, fontweight='bold', pad=15, color='#e0e0e0')
        if xlabel:
            ax.set_xlabel(xlabel, fontsize=11, color='#b0b0b0')
        if ylabel:
            ax.set_ylabel(ylabel, fontsize=11, color='#b0b0b0')
        
        ax.grid(True, alpha=0.15, linestyle='--', linewidth=0.5, color='#555')
        ax.tick_params(colors='#888', labelsize=9)
        ax.spines['top'].set_color('#333')
        ax.spines['bottom'].set_color('#333')
        ax.spines['left'].set_color('#333')
        ax.spines['right'].set_color('#333')
        
        return im
    
    def create_spectrogram(
        self,
        audio: Tensor,
        figure: plt.Figure,
        ax: plt.Axes,
        sr: int = 48000,
        n_fft: int = 1024,
        hop: int = 512,
        title: Optional[str] = None,
    ) -> PILImage.Image:
        """Create spectrogram image from audio tensor"""
        audio = torch.as_tensor(audio)
        
        w = torch.hann_window(n_fft, device=audio.device)
        spec = torch.stft(audio, n_fft, hop, window=w, return_complex=False)
        spec = spec.div_(w.pow(2).sum())
        spec = torch.view_as_complex(spec).abs().clamp_min(1e-12).log10().mul(10)
        
        vmax = max(0.0, spec.max().item())
        
        if spec.dim() > 2:
            spec = spec.squeeze(0)
        
        ax.clear()
        self.specshow(
            spec,
            ax=ax,
            title=title,
            xlabel="Time [s]",
            ylabel="Frequency [kHz]",
            sr=sr,
            n_fft=n_fft,
            hop=hop,
            vmax=vmax,
        )
        
        figure.patch.set_facecolor('#0a0a0a')
        ax.set_facecolor('#0a0a0a')
        figure.canvas.draw()
        
        return PILImage.frombytes(
            "RGB", 
            figure.canvas.get_width_height(), 
            figure.canvas.tostring_rgb()
        )


class FileManager:
    """Manages temporary file cleanup"""
    
    @staticmethod
    def cleanup_tmp(filter_list: List[str] = None, hours_keep: int = 2, temp_dir: str = "/tmp"):
        """Clean up old temporary files."""
        if filter_list is None:
            filter_list = []
        filter_list.append("p232")
        
        if not os.path.exists(temp_dir):
            return
            
        logger.info(f"Cleaning up temporary files older than {hours_keep} hours")
        cleaned = 0
        
        for filepath in glob.glob(os.path.join(temp_dir, "*")):
            try:
                is_old = (time.time() - os.path.getmtime(filepath)) / 3600 > hours_keep
                filtered = any(filt in filepath for filt in filter_list if filt is not None)
                
                if is_old and not filtered:
                    os.remove(filepath)
                    cleaned += 1
                    logger.debug(f"Removed file {filepath}")
            except Exception as e:
                logger.warning(f"Failed to remove file {filepath}: {e}")
        
        if cleaned > 0:
            logger.info(f"Cleaned up {cleaned} temporary files")


# ============================================================================
# Initialize Application
# ============================================================================

app_config = AppConfig(
    device=torch.device("cuda" if torch.cuda.is_available() else "cpu")
)

logger.info(f"Loading DeepFilterNet2 model on {app_config.device}")
model, df, _ = init_df(app_config.model_path, config_allow_defaults=True)
model = model.to(device=app_config.device).eval()

audio_processor = AudioProcessor(model, df, app_config)
audio_loader = AudioLoader()
visualizer = SpectrogramVisualizer()
file_manager = FileManager()

NOISES = {
    "None": None,
    "🍳 Kitchen": "samples/dkitchen.wav",
    "🛋️ Living Room": "samples/dliving.wav",
    "🌊 River": "samples/nriver.wav",
    "☕ Cafe": "samples/scafe.wav",
}


# ============================================================================
# Main Processing Function
# ============================================================================

def process_audio(
    speech_file: Optional[str],
    noise_type: str,
    snr: int,
    mic_input: Optional[str] = None,
) -> Tuple[str, PILImage.Image, str, PILImage.Image]:
    """Main audio processing pipeline."""
    try:
        if mic_input:
            speech_file = mic_input
        
        sr = app_config.sample_rate
        logger.info(f"Processing: file={speech_file}, noise={noise_type}, snr={snr}")
        
        if speech_file is not None:
            speech_file = audio_loader.ensure_wav(speech_file)
            sample, meta = load_audio(speech_file, sr)
            
            max_len = app_config.max_duration_seconds * sr
            if sample.shape[-1] > max_len:
                logger.warning(f"Audio too long, truncating to {app_config.max_duration_seconds}s")
                start = torch.randint(0, sample.shape[-1] - max_len, ()).item()
                sample = sample[..., start : start + max_len]
        else:
            sample, meta = load_audio("samples/p232_013_clean.wav", sr)
            sample = sample[..., : app_config.max_duration_seconds * sr]
        
        if sample.dim() > 1 and sample.shape[0] > 1:
            logger.info(f"Converting from {sample.shape[0]} channels to mono")
            sample = sample.mean(dim=0, keepdim=True)
        
        logger.info(f"Loaded audio with shape {sample.shape}")
        
        noise_fn = NOISES.get(noise_type)
        if noise_fn is not None:
            noise, _ = load_audio(noise_fn, sr)
            logger.info(f"Adding {noise_type} noise at {snr} dB SNR")
            _, _, sample = audio_processor.mix_at_snr(sample, noise, int(snr))
        
        enhanced = audio_processor.enhance_audio(sample)
        logger.info("Audio enhancement completed")
        
        if meta.sample_rate != sr and meta.sample_rate > 0:
            enhanced = resample(enhanced, sr, meta.sample_rate)
            sample = resample(sample, sr, meta.sample_rate)
            sr = meta.sample_rate
        
        noisy_wav = tempfile.NamedTemporaryFile(suffix="_noisy.wav", delete=False).name
        save_audio(noisy_wav, sample, sr)
        
        enhanced_wav = tempfile.NamedTemporaryFile(suffix="_enhanced.wav", delete=False).name
        save_audio(enhanced_wav, enhanced, sr)
        
        logger.info(f"Saved outputs: {noisy_wav}, {enhanced_wav}")
        
        noisy_spec = visualizer.create_spectrogram(
            sample, 
            visualizer.fig_noisy, 
            visualizer.ax_noisy,
            sr=sr,
            title="Input Audio Spectrogram"
        )
        
        enhanced_spec = visualizer.create_spectrogram(
            enhanced,
            visualizer.fig_enh,
            visualizer.ax_enh,
            sr=sr,
            title="Enhanced Audio Spectrogram"
        )
        
        filter_files = [speech_file, noisy_wav, enhanced_wav]
        if mic_input:
            filter_files.append(mic_input)
        file_manager.cleanup_tmp(filter_files, app_config.cleanup_hours)
        
        return noisy_wav, noisy_spec, enhanced_wav, enhanced_spec
        
    except Exception as e:
        logger.error(f"Error processing audio: {e}", exc_info=True)
        raise gr.Error(f"Processing failed: {str(e)}")


def toggle_input_mode(choice: str):
    """Toggle between microphone and file upload."""
    if choice == "mic":
        return gr.update(visible=True, value=None), gr.update(visible=False, value=None)
    else:
        return gr.update(visible=False, value=None), gr.update(visible=True, value=None)


# ============================================================================
# Custom CSS - Dark Theme
# ============================================================================

custom_css = """
/* Global Dark Theme */
.gradio-container {
    font-family: 'Inter', -apple-system, BlinkMacSystemFont, sans-serif !important;
    background: linear-gradient(135deg, #0a0a0a 0%, #1a1a2e 100%) !important;
}

body {
    background: #0a0a0a !important;
}

/* Hero Section */
#hero-section {
    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
    padding: 50px 30px;
    border-radius: 20px;
    margin-bottom: 40px;
    box-shadow: 0 15px 40px rgba(102, 126, 234, 0.6);
    text-align: center;
    border: 1px solid rgba(255, 255, 255, 0.1);
}

#hero-section h1 {
    color: #ffffff;
    font-size: 3.2em;
    font-weight: 800;
    margin: 0 0 15px 0;
    text-shadow: 2px 2px 8px rgba(0,0,0,0.4);
    letter-spacing: -1px;
}

#hero-section p {
    color: rgba(255,255,255,0.95);
    font-size: 1.25em;
    margin: 10px auto;
    max-width: 800px;
    line-height: 1.6;
    font-weight: 300;
}

/* Feature Cards - Dark */
.feature-card {
    background: linear-gradient(135deg, #1e1e2e 0%, #2d2d44 100%);
    padding: 25px;
    border-radius: 15px;
    box-shadow: 0 8px 25px rgba(0,0,0,0.4);
    margin-bottom: 20px;
    border: 1px solid rgba(102, 126, 234, 0.3);
    transition: all 0.3s ease;
}

.feature-card:hover {
    transform: translateY(-3px);
    box-shadow: 0 12px 35px rgba(102, 126, 234, 0.5);
    border-color: rgba(102, 126, 234, 0.6);
}

/* Input Controls Section */
.input-controls {
    background: linear-gradient(135deg, #1a1a2e 0%, #252545 100%);
    padding: 30px;
    border-radius: 15px;
    box-shadow: 0 5px 20px rgba(0,0,0,0.5);
    border: 1px solid rgba(102, 126, 234, 0.2);
}

/* Output Section */
.output-section {
    background: linear-gradient(135deg, #2d1b3d 0%, #3d2952 100%);
    padding: 30px;
    border-radius: 15px;
    box-shadow: 0 5px 20px rgba(0,0,0,0.5);
    border: 1px solid rgba(118, 75, 162, 0.3);
}

/* Section Headers */
.section-header {
    color: #a78bfa;
    font-size: 1.8em;
    font-weight: 700;
    margin: 30px 0 20px 0;
    text-align: center;
    text-shadow: 0 0 20px rgba(167, 139, 250, 0.5);
}

/* Process Button */
.process-button {
    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
    border: none !important;
    font-size: 1.4em !important;
    font-weight: 700 !important;
    padding: 20px 50px !important;
    border-radius: 50px !important;
    box-shadow: 0 10px 40px rgba(102, 126, 234, 0.7) !important;
    transition: all 0.3s ease !important;
    color: #ffffff !important;
    text-transform: uppercase;
    letter-spacing: 1px;
}

.process-button:hover {
    transform: translateY(-3px) scale(1.02) !important;
    box-shadow: 0 15px 50px rgba(102, 126, 234, 0.9) !important;
}

/* Audio Components */
.audio-wrapper {
    background: linear-gradient(135deg, #1e1e2e 0%, #2a2a40 100%);
    padding: 20px;
    border-radius: 12px;
    box-shadow: 0 3px 12px rgba(0,0,0,0.6);
    margin: 15px 0;
    border: 1px solid rgba(102, 126, 234, 0.2);
}

/* Tabs */
.tab-nav button {
    font-weight: 600 !important;
    font-size: 1.1em !important;
    padding: 12px 24px !important;
    border-radius: 10px 10px 0 0 !important;
    background: #1a1a2e !important;
    color: #a0a0b0 !important;
    border: 1px solid rgba(102, 126, 234, 0.2) !important;
}

.tab-nav button[aria-selected="true"] {
    background: linear-gradient(135deg, #667eea, #764ba2) !important;
    color: #ffffff !important;
}

/* Info Box */
.info-box {
    background: linear-gradient(135deg, #1e1e3f 0%, #2d2d52 100%);
    padding: 25px;
    border-radius: 15px;
    margin: 25px 0;
    border-left: 5px solid #667eea;
    box-shadow: 0 4px 20px rgba(0,0,0,0.5);
}

.info-box h3 {
    color: #a78bfa;
    font-size: 1.4em;
    font-weight: 700;
    margin-top: 0;
}

.info-box p, .info-box ul, .info-box li {
    color: #c0c0d0;
}

.info-box ul {
    margin: 10px 0;
    padding-left: 25px;
}

.info-box li {
    margin: 8px 0;
    line-height: 1.6;
}

/* Examples Section */
.examples-section {
    background: linear-gradient(135deg, #2a2a3e 0%, #35354f 100%);
    padding: 25px;
    border-radius: 15px;
    margin-top: 30px;
    box-shadow: 0 4px 20px rgba(0,0,0,0.5);
    border: 1px solid rgba(102, 126, 234, 0.2);
}

/* Footer */
#footer {
    text-align: center;
    padding: 30px 20px;
    margin-top: 50px;
    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
    border-radius: 15px;
    color: #ffffff;
    box-shadow: 0 10px 30px rgba(102, 126, 234, 0.5);
}

#footer h3 {
    margin: 0 0 10px 0;
    font-size: 1.5em;
    font-weight: 700;
}

#footer p {
    margin: 5px 0;
    opacity: 0.95;
}

/* Radio Buttons */
.radio-group label {
    padding: 12px 20px !important;
    border-radius: 10px !important;
    font-weight: 600 !important;
    transition: all 0.3s ease !important;
    background: #1a1a2e !important;
    color: #b0b0c0 !important;
    border: 1px solid rgba(102, 126, 234, 0.3) !important;
}

.radio-group label:hover {
    background: #252545 !important;
    border-color: rgba(102, 126, 234, 0.6) !important;
}

/* Dropdowns */
.dropdown select {
    border-radius: 10px !important;
    padding: 12px !important;
    font-size: 1.05em !important;
    background: #1a1a2e !important;
    color: #c0c0d0 !important;
    border: 2px solid rgba(102, 126, 234, 0.3) !important;
    transition: all 0.3s ease !important;
}

.dropdown select:focus {
    border-color: #667eea !important;
    box-shadow: 0 0 0 3px rgba(102, 126, 234, 0.3) !important;
}

/* Labels and Text */
label, .label {
    color: #b0b0c0 !important;
}

/* Markdown Text */
.markdown-text, .prose {
    color: #c0c0d0 !important;
}

/* Input Fields */
input, textarea {
    background: #1a1a2e !important;
    color: #c0c0d0 !important;
    border: 1px solid rgba(102, 126, 234, 0.3) !important;
}

/* Scrollbars */
::-webkit-scrollbar {
    width: 10px;
    background: #1a1a2e;
}

::-webkit-scrollbar-thumb {
    background: linear-gradient(135deg, #667eea, #764ba2);
    border-radius: 5px;
}

::-webkit-scrollbar-thumb:hover {
    background: linear-gradient(135deg, #764ba2, #667eea);
}
"""

# ============================================================================
# Gradio Interface
# ============================================================================

with gr.Blocks(css=custom_css, theme=gr.themes.Base()) as demo:
    
    # Hero Section
    gr.HTML("""
        <div id="hero-section">
            <h1>🎵 DeepFilterNet2 Audio Enhancement</h1>
            <p>Transform noisy audio into crystal-clear sound using cutting-edge AI technology</p>
            <p style="font-size: 0.95em; margin-top: 15px;">
                ✨ Real-time Processing | 🎯 State-of-the-Art Quality | 🚀 Lightning Fast
            </p>
        </div>
    """)
    
    # Quick Start Guide
    with gr.Row():
        gr.Markdown("""
        <div class="info-box">
            <h3>🚀 Quick Start Guide</h3>
            <ul>
                <li><strong>Step 1:</strong> Upload an audio file or record using your microphone</li>
                <li><strong>Step 2:</strong> Optionally add synthetic noise to test the denoiser</li>
                <li><strong>Step 3:</strong> Adjust SNR settings if needed</li>
                <li><strong>Step 4:</strong> Click the "Denoise Audio" button</li>
                <li><strong>Step 5:</strong> Compare results with interactive spectrograms</li>
            </ul>
        </div>
        """)
    
    # Main Interface
    with gr.Row():
        # Left Column - Input Controls
        with gr.Column(scale=1):
            gr.HTML('<h2 class="section-header">📤 Audio Input</h2>')
            
            with gr.Group(elem_classes="input-controls"):
                input_mode = gr.Radio(
                    ["file", "mic"],
                    value="file",
                    label="🎙️ Input Method",
                    info="Choose your preferred input source",
                    elem_classes="radio-group"
                )
                
                audio_file = gr.Audio(
                    type="filepath",
                    label="📁 Upload Audio File",
                    visible=True,
                    elem_classes="audio-wrapper"
                )
                
                mic_input = gr.Audio(
                    sources=["microphone"],
                    type="filepath",
                    label="🎤 Record Audio",
                    visible=False,
                    elem_classes="audio-wrapper"
                )
            
            gr.HTML('<h2 class="section-header">⚙️ Enhancement Settings</h2>')
            
            with gr.Group(elem_classes="feature-card"):
                noise_type = gr.Dropdown(
                    label="🔊 Background Noise Type",
                    choices=list(NOISES.keys()),
                    value="None",
                    info="Add synthetic noise for testing",
                    elem_classes="dropdown"
                )
                
                snr = gr.Dropdown(
                    label="📊 Signal-to-Noise Ratio (dB)",
                    choices=["-5", "0", "10", "20"],
                    value="10",
                    info="Higher = cleaner signal",
                    elem_classes="dropdown"
                )
            

            
            process_btn = gr.Button(
                "🚀 Denoise Audio",
                elem_classes="process-button",
                size="lg"
            )
            
        # Right Column - Results
        with gr.Column(scale=2):
            gr.HTML('<h2 class="section-header">📊 Results & Comparison</h2>')
            
            with gr.Tabs():
                with gr.Tab("🔴 Input Audio", elem_classes="output-section"):
                    noisy_audio = gr.Audio(
                        type="filepath",
                        label="Original/Noisy Audio",
                        elem_classes="audio-wrapper"
                    )
                    noisy_spec = gr.Image(
                        label="Input Spectrogram",
                        elem_classes="audio-wrapper"
                    )
                
                with gr.Tab("🟢 Enhanced Audio", elem_classes="output-section"):
                    enhanced_audio = gr.Audio(
                        type="filepath",
                        label="Enhanced Audio",
                        elem_classes="audio-wrapper"
                    )
                    enhanced_spec = gr.Image(
                        label="Enhanced Spectrogram",
                        elem_classes="audio-wrapper"
                    )
    
    # Examples Section
    gr.HTML('<h2 class="section-header">🎯 Try These Examples</h2>')
    
    with gr.Group(elem_classes="examples-section"):
        gr.Examples(
            examples=[
                ["./samples/p232_013_clean.wav", "🍳 Kitchen", "10"],
                ["./samples/p232_013_clean.wav", "☕ Cafe", "10"],
                ["./samples/p232_019_clean.wav", "☕ Cafe", "10"],
                ["./samples/p232_019_clean.wav", "🌊 River", "10"],
            ],
            inputs=[audio_file, noise_type, snr],
            outputs=[noisy_audio, noisy_spec, enhanced_audio, enhanced_spec],
            fn=process_audio,
            cache_examples=True,
            label="Click any example to try it instantly",
        )
    
    # Technical Information
    with gr.Row():
        with gr.Column():
            gr.Markdown("""
            <div class="info-box">
                <h3>💡 How It Works</h3>
                <p><strong>DeepFilterNet2</strong> uses advanced deep learning to identify and remove unwanted background noise while preserving speech clarity. The model analyzes spectral patterns to distinguish between signal and noise components.</p>
            </div>
            """)
        
        with gr.Column():
            gr.Markdown("""
            <div class="info-box">
                <h3>📋 Technical Specifications</h3>
                <ul>
                    <li><strong>Model:</strong> DeepFilterNet2 (State-of-the-art)</li>
                    <li><strong>Sample Rate:</strong> 48 kHz</li>
                    <li><strong>Max Duration:</strong> 1 hour</li>
                    <li><strong>Formats:</strong> WAV, MP3, M4A, OGG, FLAC, AAC</li>
                    <li><strong>Processing:</strong> Real-time capable</li>
                </ul>
            </div>
            """)
    
    # Footer
    gr.HTML("""
        <div id="footer">
            <h3>🎵 Powered by DeepFilterNet2</h3>
            <p>Advanced AI-driven audio enhancement technology</p>
            <p><em>Built with Gradio • Optimized for Performance</em></p>
        </div>
    """)
    
    # Event Handlers
    process_btn.click(
        fn=process_audio,
        inputs=[audio_file, noise_type, snr, mic_input],
        outputs=[noisy_audio, noisy_spec, enhanced_audio, enhanced_spec],
        api_name="denoise",
    )
    
    input_mode.change(
        fn=toggle_input_mode,
        inputs=input_mode,
        outputs=[mic_input, audio_file],
    )

# Initial cleanup
file_manager.cleanup_tmp()

# Launch application
if __name__ == "__main__":
    demo.queue().launch(
        server_name="0.0.0.0",
        server_port=7860,
        share=False,
    )