Spaces:
Runtime error
Runtime error
File size: 2,285 Bytes
5d5905f a17627f 5d5905f a17627f 5d5905f a17627f 5d5905f a17627f 5d5905f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 |
import gradio as gr
from speechbrain.inference.ASR import EncoderASR
from pydub import AudioSegment
import os
import numpy as np
from scipy.io.wavfile import write
# Load the ASR model
asr_model = EncoderASR.from_hparams(
source="YosefA/wave2vec2_amharic_stt",
savedir="pretrained_models/asr-wav2vec2-amharic"
)
# Directory to store converted audio files
os.makedirs("temp_audio", exist_ok=True)
def transcribe_audio(audio_file):
"""
Converts the uploaded audio to .wav format, runs transcription, and returns the result.
"""
# Save the uploaded audio to a temporary location
temp_audio_path = "temp_audio/input_audio.wav"
# Convert audio to .wav format
sound = AudioSegment.from_file(audio_file)
sound.export(temp_audio_path, format="wav")
# Transcribe the audio
transcription = asr_model.transcribe_file(temp_audio_path)
# Clean up temporary files (optional)
os.remove(temp_audio_path)
return transcription
def save_audio_to_file(audio_data, file_path="temp_audio/input_audio.wav"):
"""
Converts Gradio audio data (numpy array and sample rate) to a .wav file.
"""
audio_array, sample_rate = audio_data
# Normalize and save as .wav
os.makedirs(os.path.dirname(file_path), exist_ok=True)
write(file_path, sample_rate, np.int16(audio_array * 32767)) # Convert float32 to int16
return file_path
def process_audio(audio_data):
"""
Processes recorded/uploaded audio, saves it, and sends it to the transcribe_audio function.
"""
temp_audio_path = save_audio_to_file(audio_data)
transcription = transcribe_audio(temp_audio_path)
return transcription
# Define the Gradio interface
with gr.Blocks() as app:
gr.Markdown("### Amharic Speech-to-Text Transcription App")
gr.Markdown("Upload or record an audio file in any format, and get its transcription.")
with gr.Row():
audio_input = gr.Audio(label="Upload or Record Audio", type="filepath")
transcription_output = gr.Textbox(label="Transcription")
transcribe_button = gr.Button("Transcribe")
transcribe_button.click(transcribe_audio, inputs=audio_input, outputs=transcription_output)
# Launch the app
if __name__ == "__main__":
app.launch()
|