added app for automatic speech recognition

app.py

@@ -1,7 +1,16 @@
 import gradio as gr
 
-def greet(name):
-    return "Hello " + name + "!!"
+from tasks.asr import (
+    mic_transcribe_interface,
+    file_transcribe_interface
+)
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+app = gr.Blocks()
+
+with app:
+    gr.TabbedInterface(
+        [mic_transcribe_interface, file_transcribe_interface],
+        ["Transcribe from Microphone", "Transcribe from Audio File"]
+    )
+
+app.launch(share=True)

requirements.txt

@@ -0,0 +1,8 @@
+torch
+librosa
+soundfile
+transformers 
+pillow
+numpy
+requests
+matplotlib

tasks/asr.py

@@ -0,0 +1,47 @@
+import librosa 
+import numpy as np
+import gradio as gr 
+import soundfile as sf 
+from transformers import pipeline 
+
+# Load the pipeline
+model = pipeline(
+    task="automatic-speech-recognition",
+    model="distil-whisper/distil-small.en"
+)
+
+def transcribe_audio(filepath):
+    """Transcribe audio to text"""
+    audio, sample_rate = sf.read(filepath)
+    audio_mono = librosa.to_mono(np.transpose(audio))
+    
+    # resample the audio
+    audio_16KHz = librosa.resample(
+        audio_mono,
+        orig_sr=sample_rate,
+        target_sr=16000
+    )
+
+    output = model(
+        audio_16KHz,
+        chunk_length_s=30,
+        batch_size=4,
+    )
+
+    return output["text"]
+
+mic_transcribe_interface = gr.Interface(
+    fn=transcribe_audio,
+    inputs=gr.Audio(sources="microphone", type="filepath"),
+    outputs=gr.Textbox(label="Transcription", lines=3),
+    allow_flagging="never",
+    title="Transcribe Audio from your Microphone"
+)
+
+file_transcribe_interface = gr.Interface(
+    fn=transcribe_audio,
+    inputs=gr.Audio(sources="upload", type="filepath"),
+    outputs=gr.Textbox(label="Transcription", lines=3),
+    allow_flagging="never",
+    title="Transcribe Audio from a File"
+)