Spaces:
Running
Running
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import pyttsx3
|
| 3 |
+
import os
|
| 4 |
+
|
| 5 |
+
# Initialize the TTS engine
|
| 6 |
+
engine = pyttsx3.init()
|
| 7 |
+
|
| 8 |
+
# Get available voices
|
| 9 |
+
voices = engine.getProperty('voices')
|
| 10 |
+
|
| 11 |
+
# Create a dictionary to map voice names to IDs
|
| 12 |
+
voice_map = {f"{voice.name} ({voice.languages[0] if voice.languages else 'Unknown'})": voice.id for voice in voices}
|
| 13 |
+
|
| 14 |
+
def text_to_speech(text, voice_name, rate=200):
|
| 15 |
+
"""
|
| 16 |
+
Convert text to speech with selected voice and rate.
|
| 17 |
+
Saves output as an audio file and returns the file path.
|
| 18 |
+
"""
|
| 19 |
+
# Initialize engine for each call to avoid threading issues with Gradio
|
| 20 |
+
engine = pyttsx3.init()
|
| 21 |
+
|
| 22 |
+
# Set voice
|
| 23 |
+
voice_id = voice_map.get(voice_name)
|
| 24 |
+
if not voice_id:
|
| 25 |
+
return "Error: Selected voice not found."
|
| 26 |
+
engine.setProperty('voice', voice_id)
|
| 27 |
+
|
| 28 |
+
# Set speech rate
|
| 29 |
+
engine.setProperty('rate', rate)
|
| 30 |
+
|
| 31 |
+
# Save audio to a file
|
| 32 |
+
output_file = "output.wav"
|
| 33 |
+
engine.save_to_file(text, output_file)
|
| 34 |
+
engine.runAndWait()
|
| 35 |
+
|
| 36 |
+
return output_file
|
| 37 |
+
|
| 38 |
+
# Gradio interface
|
| 39 |
+
with gr.Blocks(title="Text-to-Speech with Different Voices") as demo:
|
| 40 |
+
gr.Markdown("# Text-to-Speech Converter")
|
| 41 |
+
gr.Markdown("Enter text and select a voice to convert it to speech with different voices and accents.")
|
| 42 |
+
|
| 43 |
+
text_input = gr.Textbox(label="Enter Text", placeholder="Type your text here...")
|
| 44 |
+
voice_dropdown = gr.Dropdown(choices=list(voice_map.keys()), label="Select Voice/Accent")
|
| 45 |
+
rate_slider = gr.Slider(minimum=100, maximum=300, value=200, step=10, label="Speech Rate")
|
| 46 |
+
|
| 47 |
+
convert_button = gr.Button("Convert to Speech")
|
| 48 |
+
audio_output = gr.Audio(label="Generated Speech")
|
| 49 |
+
|
| 50 |
+
convert_button.click(
|
| 51 |
+
fn=text_to_speech,
|
| 52 |
+
inputs=[text_input, voice_dropdown, rate_slider],
|
| 53 |
+
outputs=audio_output
|
| 54 |
+
)
|
| 55 |
+
|
| 56 |
+
# Launch the app (commented out for Hugging Face Spaces deployment)
|
| 57 |
+
# demo.launch()
|
| 58 |
+
|
| 59 |
+
if __name__ == "__main__":
|
| 60 |
+
demo.launch()
|