Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import speech_recognition as sr
|
2 |
+
from pydub import AudioSegment
|
3 |
+
import gradio as gr
|
4 |
+
from os import path
|
5 |
+
import requests
|
6 |
+
import openai
|
7 |
+
|
8 |
+
r = sr.Recognizer()
|
9 |
+
|
10 |
+
def record_text(audio_file):
|
11 |
+
sound = audio_file
|
12 |
+
sound_type = sound.split(".")
|
13 |
+
if sound_type[-1] == 'mp3':
|
14 |
+
input_file = sound
|
15 |
+
output_file = "con_sound.wav"
|
16 |
+
try:
|
17 |
+
# convert mp3 file to wav file
|
18 |
+
sound = AudioSegment.from_mp3(input_file)
|
19 |
+
sound.export(output_file, format="wav")
|
20 |
+
sound = "con_sound.wav"
|
21 |
+
except Exception as e:
|
22 |
+
print(f"Error converting MP3 to WAV: {e}")
|
23 |
+
return ""
|
24 |
+
|
25 |
+
MyText = ""
|
26 |
+
with sr.AudioFile(sound) as source:
|
27 |
+
r.adjust_for_ambient_noise(source)
|
28 |
+
print("Converting audio file to text..")
|
29 |
+
audio2 = r.record(source, duration=None) # Use record instead of listen
|
30 |
+
try:
|
31 |
+
MyText = r.recognize_google(audio2, language="en-US", key=None, show_all=False)
|
32 |
+
MyText = MyText.lower()
|
33 |
+
print("Converted audio is:\n" + MyText + '.')
|
34 |
+
except sr.UnknownValueError:
|
35 |
+
print("Google Speech Recognition could not understand audio")
|
36 |
+
|
37 |
+
return MyText
|
38 |
+
|
39 |
+
def message_and_history(audio_file, history, api_key):
|
40 |
+
history = history or []
|
41 |
+
input_text = "Type and press Enter"
|
42 |
+
output_text = record_text(audio_file)
|
43 |
+
|
44 |
+
if len(input_text) == 0:
|
45 |
+
input_text = "Apply proper punctuation on the given paragraph."
|
46 |
+
history.append((input_text, output_text))
|
47 |
+
else:
|
48 |
+
history.append((input_text, output_text))
|
49 |
+
|
50 |
+
return history, history
|
51 |
+
|
52 |
+
prompt = "Type and press Enter"
|
53 |
+
block = gr.Blocks(theme=gr.themes.Glass(primary_hue="slate"))
|
54 |
+
with block:
|
55 |
+
gr.Markdown("""<h1><center>Audio Recognition - Ask and Learn about an Audio</center></h1> """)
|
56 |
+
with gr.Row():
|
57 |
+
with gr.Column(scale=0.5):
|
58 |
+
aud_input = gr.Audio(type="filepath", label="Upload Audio")
|
59 |
+
api_input = gr.Textbox(label="Enter Api-key")
|
60 |
+
upload_button = gr.Button(value="Upload & Start Chat", interactive=True, variant="primary")
|
61 |
+
with gr.Column():
|
62 |
+
chatbot = gr.Chatbot(label="Ask questions about the audio")
|
63 |
+
message = gr.Textbox(label="User", placeholder=prompt)
|
64 |
+
state = gr.State()
|
65 |
+
|
66 |
+
upload_button.click(message_and_history, inputs=[aud_input, state, api_input], outputs=[chatbot, state])
|
67 |
+
message.submit(message_and_history, inputs=[aud_input, state, api_input], outputs=[chatbot, state])
|
68 |
+
message.submit(lambda: None, None, message, queue=False)
|
69 |
+
block.launch(share=True)
|