Commit
·
29b5120
1
Parent(s):
1d62a43
Simplifying dependencies in audio part
Browse files- app.py +13 -5
- backend/audio_to_tgt.py +5 -7
app.py
CHANGED
|
@@ -62,12 +62,20 @@ description_audio = "Upload an audio file to extract text and translate it to En
|
|
| 62 |
|
| 63 |
audio_interface = gr.Interface(
|
| 64 |
fn=src_audio_to_eng_translator,
|
| 65 |
-
inputs=[
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 66 |
gr.Dropdown(
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
],
|
| 72 |
outputs=[gr.Textbox(label="Original text"),
|
| 73 |
gr.Textbox(label="Translated text"),
|
|
|
|
| 62 |
|
| 63 |
audio_interface = gr.Interface(
|
| 64 |
fn=src_audio_to_eng_translator,
|
| 65 |
+
inputs=[
|
| 66 |
+
gr.Audio(
|
| 67 |
+
label="Upload an Audio file",
|
| 68 |
+
type="filepath"
|
| 69 |
+
),
|
| 70 |
+
gr.Dropdown(
|
| 71 |
+
choices=["turbo", "base", "tiny", "small", "medium", "large"],
|
| 72 |
+
label="Select Whisper Model size",
|
| 73 |
+
),
|
| 74 |
gr.Dropdown(
|
| 75 |
+
choices=language_list,
|
| 76 |
+
label="Select Target Language",
|
| 77 |
+
interactive=True
|
| 78 |
+
)
|
| 79 |
],
|
| 80 |
outputs=[gr.Textbox(label="Original text"),
|
| 81 |
gr.Textbox(label="Translated text"),
|
backend/audio_to_tgt.py
CHANGED
|
@@ -1,7 +1,6 @@
|
|
| 1 |
import whisper
|
| 2 |
import numpy as np
|
| 3 |
from pydub import AudioSegment
|
| 4 |
-
import langcodes
|
| 5 |
from deep_translator import GoogleTranslator, detection
|
| 6 |
import os
|
| 7 |
|
|
@@ -24,10 +23,9 @@ def src_audio_to_eng_translator(audio_file_input, model_size = "turbo", target_l
|
|
| 24 |
model = whisper.load_model(model_size)
|
| 25 |
result = model.transcribe(audio_data)
|
| 26 |
input_text = result["text"]
|
| 27 |
-
|
| 28 |
src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
|
| 29 |
-
src_lang =
|
| 30 |
-
|
| 31 |
-
translated_text = GoogleTranslator(source='auto', target=
|
| 32 |
-
return input_text, translated_text, src_lang
|
| 33 |
-
# return result['text']
|
|
|
|
| 1 |
import whisper
|
| 2 |
import numpy as np
|
| 3 |
from pydub import AudioSegment
|
|
|
|
| 4 |
from deep_translator import GoogleTranslator, detection
|
| 5 |
import os
|
| 6 |
|
|
|
|
| 23 |
model = whisper.load_model(model_size)
|
| 24 |
result = model.transcribe(audio_data)
|
| 25 |
input_text = result["text"]
|
| 26 |
+
|
| 27 |
src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
|
| 28 |
+
src_lang = formatted_languages.get(src_lang_code, 'Source language not detected')
|
| 29 |
+
target_lang_code = formatted_languages.get(target_lang, 'en')
|
| 30 |
+
translated_text = GoogleTranslator(source='auto', target=target_lang_code).translate(input_text)
|
| 31 |
+
return input_text, translated_text, src_lang
|
|
|