import gradio as gr
import argparse
import os
import torch
from time import sleep
from tqdm import tqdm
from lang_list import union_language_dict
# import pyperclip
from pytube import YouTube
import re
from PIL import Image
import urllib.request
NUMBER = 100
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
DEVICE = "cpu"
DOWNLOAD = True
SLICE_AUDIO = True
TRANSCRIBE_AUDIO = True
CONCATENATE_TRANSCRIPTIONS = True
TRANSLATE_TRANSCRIPTIONS = True
ADD_SUBTITLES_TO_VIDEO = True
REMOVE_FILES = True
if DEVICE == "cpu":
# I supose that I am on huggingface server
# Get RAM space
ram = int(os.popen("free -m | grep Mem | awk '{print $2}'").read())
factor = 1
SECONDS = int(ram*factor)
print(f"RAM: {ram}, SECONDS: {SECONDS}")
else:
# I supose that I am on my computer
# Get VRAM space
SECONDS = 300
YOUTUBE = "youtube"
TWITCH = "twitch"
ERROR = "error"
urllib.request.urlretrieve('https://maximofn.com/wp-content/uploads/2023/11/subtify_logo-scaled.webp', "subtify_logo.webp")
subtify_logo = Image.open("subtify_logo.webp")
subtify_logo_width, subtify_logo_height = subtify_logo.size
factor = 4
new_width = subtify_logo_width // factor
new_height = subtify_logo_height // factor
BACKGROUND_COLOR = "#0b0f19"
BUTTON_COLOR = "#47515f"
SVG_COLOR = "#f3f4f6"
PANEL_COLOR = "#101827"
PRIMARY_TEXT_COLOR = "#f3f4f6"
SUBDUED_TEXT_COLOR = "#59616f"
BACKGROUND_PRIMARY_COLOR = "#1f2937"
BACKGROUND_SECONDARY_COLOR = "#101827"
PRIMARY_BODER_COLOR = "#323c4c"
BLOCK_TITLE_TEXT_COLOR = "#dfe2e6"
INPUT_BACKGROUND_COLOR = "#2f3947"
INPUT_BORDER_COLOR = "#313b4b"
INPUT_PLACEHOLDER_COLOR = "#616977"
ERROR_BACKGROUND_COLOR = "#101827"
ERROR_TEXT_COLOR = "#f7f2f2"
ERROR_BORDER_COLOR = "#9b3339"
BUTTON_SECONDARY_BACKGROUND_COLOR = "#434d5c"
BUTTON_SECONDARY_BORDER_COLOR = "#444d5b"
BUTTON_SECONDARY_TEXT_COLOR = "#c5c9cc"
RED = "#ff0000"
GREEN = "#00ff00"
BLUE = "#0000ff"
html_social_media = f'''
'''
html_subtify_logo = f"""
"""
html_buy_me_a_coffe = '''
'''
language_dict = union_language_dict()
def subtify_no_ui():
number_works = 6
progress_bar = tqdm(total=number_works, desc="Subtify")
folder_chunck = "chunks"
folder_concatenated = "concatenated_transcriptions"
folder_translated_transcriptions = "translated_transcriptions"
if not os.path.exists(folder_chunck):
os.makedirs(folder_chunck)
if not os.path.exists(folder_concatenated):
os.makedirs(folder_concatenated)
if not os.path.exists(folder_translated_transcriptions):
os.makedirs(folder_translated_transcriptions)
################## Download video and audio ##################
if DOWNLOAD:
print('*'*NUMBER)
# url = "https://www.twitch.tv/videos/1936119752" # twitch Rob Mula 2 horas
# url = "https://www.youtube.com/watch?v=yX5EJf4R77s" # ✅ debate, varios hablantes, 3 minutos
# url = "https://www.youtube.com/watch?v=cgx0QnXo1OU" # ✅ smart home, un solo hablante, 4:42 minutos
# url = "https://www.youtube.com/watch?v=dgOBxhi19T8" # ✅ rob mula, muchos hablantes, 4:28 minutos
# url = "https://www.youtube.com/watch?v=Coj72EzmX20" # rob mula, un solo hablante, 16 minutos
# url = "https://www.youtube.com/watch?v=Tqth0fKo0_g" # Conversación short
url = "https://www.youtube.com/watch?v=h9xPrgTYP_0" # Letitia 40 segundos
print(f"Downloading video and audio from {url}")
python_file = "download.py"
command = f"python {python_file} {url}"
os.system(command)
sleep(1)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
################## Slice audio ##################
if SLICE_AUDIO:
print('*'*NUMBER)
print("Slicing audio")
python_file = "slice_audio.py"
audio = "audios/download_audio.mp3"
command = f"python {python_file} {audio} {SECONDS}"
os.system(command)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
################# Transcript slices ##################
if TRANSCRIBE_AUDIO:
print('*'*NUMBER)
print("Transcript slices")
chunks_folder = "chunks"
if not os.path.exists(chunks_folder):
os.makedirs(chunks_folder)
python_file = "transcribe.py"
chunks_file = "chunks/output_files.txt"
number_of_speakers = 10
source_languaje = "English"
command = f"python {python_file} {chunks_file} {source_languaje} {number_of_speakers} {DEVICE}"
os.system(command)
if REMOVE_FILES:
with open(chunks_file, 'r') as f:
files = f.read().splitlines()
for file in files:
audios_extension = "mp3"
file_name, _ = file.split(".")
_, file_name = file_name.split("/")
vocal = f'{chunks_folder}/{file_name}.{audios_extension}'
command = f"rm {vocal}"
os.system(command)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
################## Concatenate transcriptions ##################
if CONCATENATE_TRANSCRIPTIONS:
print('*'*NUMBER)
print("Concatenate transcriptions")
folder_concatenated = "concatenated_transcriptions"
if not os.path.exists(folder_concatenated):
os.makedirs(folder_concatenated)
chunck_file = "chunks/output_files.txt"
python_file = "concat_transcriptions.py"
command = f"python {python_file} {chunck_file} {SECONDS}"
os.system(command)
if REMOVE_FILES:
with open(chunck_file, 'r') as f:
files = f.read().splitlines()
for file in files:
file_name, _ = file.split(".")
_, file_name = file_name.split("/")
transcriptions_folder = "transcriptions"
transcription_extension = "srt"
command = f"rm {transcriptions_folder}/{file_name}.{transcription_extension}"
os.system(command)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
################## Translate transcription ##################
target_languaje = "Español"
if TRANSLATE_TRANSCRIPTIONS:
print('*'*NUMBER)
print("Translate transcription")
transcription_file = "concatenated_transcriptions/download_audio.srt"
source_languaje = "English"
python_file = "translate_transcriptions.py"
command = f"python {python_file} {transcription_file} --source_languaje {source_languaje} --target_languaje {target_languaje} --device {DEVICE}"
os.system(command)
if REMOVE_FILES:
if os.path.exists(transcription_file):
command = f"rm {transcription_file}"
os.system(command)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
################## Add subtitles to video ##################
if ADD_SUBTITLES_TO_VIDEO:
print('*'*NUMBER)
print("Add subtitles to video")
python_file = "add_subtitles_to_video.py"
transcription_file = f"translated_transcriptions/download_audio_{target_languaje}.srt"
input_video_file = "videos/download_video.mp4"
input_audio_file = "audios/download_audio.mp3"
command = f"python {python_file} {transcription_file} {input_video_file} {input_audio_file}"
os.system(command)
if REMOVE_FILES:
if os.path.exists(input_video_file):
command = f"rm {input_video_file}"
os.system(command)
if os.path.exists(input_audio_file):
command = f"rm {input_audio_file}"
os.system(command)
if os.path.exists(transcription_file):
command = f"rm {transcription_file}"
os.system(command)
if os.path.exists("chunks/output_files.txt"):
command = f"rm chunks/output_files.txt"
os.system(command)
if os.path.exists("chunks"):
command = f"rm -r chunks"
os.system(command)
if os.path.exists("vocals/speakers.txt"):
command = f"rm vocals/speakers.txt"
os.system(command)
print('*'*NUMBER)
print("\n\n")
progress_bar.update(1)
def remove_all_files():
if os.path.exists("audios"):
command = f"rm -r audios"
os.system(command)
if os.path.exists("chunks"):
command = f"rm -r chunks"
os.system(command)
if os.path.exists("concatenated_transcriptions"):
command = f"rm -r concatenated_transcriptions"
os.system(command)
if os.path.exists("transcriptions"):
command = f"rm -r transcriptions"
os.system(command)
if os.path.exists("translated_transcriptions"):
command = f"rm -r translated_transcriptions"
os.system(command)
if os.path.exists("videos"):
command = f"rm -r videos"
os.system(command)
if os.path.exists("vocals"):
command = f"rm -r vocals"
os.system(command)
# def paste_url_from_clipboard():
# return pyperclip.paste()
def reset_frontend():
visible = False
return (
"",
gr.Image(visible=visible),
gr.Dropdown(visible=visible),
gr.Dropdown(visible=visible),
gr.Dropdown(visible=visible),
gr.Accordion(visible=visible),
gr.Button(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Textbox(visible=visible),
gr.Video(visible=visible),
)
def show_auxiliar_block1():
return gr.Textbox(value="URL checked", visible=False)
def get_youtube_thumbnail(url):
yt = YouTube(url)
thumbnail_url = yt.thumbnail_url
return thumbnail_url
def is_valid_youtube_url(url):
# This regular expression should match the following YouTube URL formats:
# - https://youtube.com/watch?v=video_id
# - https://www.youtube.com/watch?v=video_id
# - https://youtu.be/video_id
patron_youtube = r'(https?://)?(www\.)?(youtube\.com/watch\?v=|youtu\.be/)[\w-]+'
return bool(re.match(patron_youtube, url))
def is_valid_twitch_url(url):
# This regular expression should match the following Twitch URL formats:
# - https://twitch.tv/channel_name
# - https://www.twitch.tv/channel_name
# - https://twitch.tv/videos/video_id
twitch_pattern = r'(https?://)?(www\.)?twitch\.tv/(videos/\d+|\w+)'
return bool(re.match(twitch_pattern, url))
def is_valid_url(url):
num_speaker = [10, 9, 8, 7, 6, 5, 4, 3, 2, 1]
source_languaje = gr.Dropdown(visible=True, label="Source languaje", show_label=True, value="English", choices=language_dict, scale=1, interactive=True)
target_languaje = gr.Dropdown(visible=True, label="Target languaje", show_label=True, value="Español", choices=language_dict, scale=1, interactive=True)
advanced_setings = gr.Accordion(visible=True)
number_of_speakers = gr.Dropdown(visible=True, label="Number of speakers", show_label=True, value=10, choices=num_speaker, scale=1, interactive=True)
subtify_button = gr.Button(size="lg", value="subtify", min_width="10px", scale=0, visible=True)
# Youtube
if "youtube" in url.lower() or "youtu.be" in url.lower():
if is_valid_youtube_url(url):
thumbnail = get_youtube_thumbnail(url)
if thumbnail:
return (
gr.Image(value=thumbnail, visible=True, show_download_button=False, container=False),
source_languaje,
target_languaje,
advanced_setings,
number_of_speakers,
subtify_button,
)
else:
return (
gr.Image(value="assets/youtube-no-thumbnails.webp", visible=True, show_download_button=False, container=False),
source_languaje,
target_languaje,
advanced_setings,
number_of_speakers,
subtify_button,
)
# Twitch
elif "twitch" in url.lower() or "twitch.tv" in url.lower():
if is_valid_twitch_url(url):
return (
gr.Image(value="assets/twitch.webp", visible=True, show_download_button=False, container=False),
source_languaje,
target_languaje,
advanced_setings,
number_of_speakers,
subtify_button,
)
# Error
visible = False
image = gr.Image(value="assets/youtube_error.webp", visible=visible, show_download_button=False, container=False)
source_languaje = gr.Dropdown(visible=visible, label="Source languaje", show_label=True, value="English", choices=language_dict, scale=1, interactive=True)
target_languaje = gr.Dropdown(visible=visible, label="Target languaje", show_label=True, value="Español", choices=language_dict, scale=1, interactive=True)
advanced_setings = gr.Accordion(visible=visible)
number_of_speakers = gr.Dropdown(visible=visible, label="Number of speakers", show_label=True, value=10, choices=num_speaker, scale=1, interactive=True)
subtify_button = gr.Button(size="lg", value="subtify", min_width="10px", scale=0, visible=visible)
return (
image,
source_languaje,
target_languaje,
advanced_setings,
number_of_speakers,
subtify_button,
)
def change_visibility_texboxes():
return (
gr.Textbox(value="Done"),
gr.Textbox(visible=True),
gr.Textbox(visible=True),
gr.Textbox(visible=True),
gr.Textbox(visible=True),
gr.Textbox(visible=True),
gr.Textbox(visible=True),
gr.Textbox(visible=False),
)
def get_audio_and_video_from_video(url):
print('*'*NUMBER)
print(f"Downloading video and audio from {url}")
audios_folder = "audios"
videos_folder = "videos"
if not os.path.exists(audios_folder):
os.makedirs(audios_folder)
if not os.path.exists(videos_folder):
os.makedirs(videos_folder)
python_file = "download.py"
command = f"python {python_file} {url}"
os.system(command)
sleep(1)
audio = "audios/download_audio.mp3"
video = "videos/download_video.mp4"
return (
gr.Textbox(value="Ok"),
gr.Textbox(value=audio),
gr.Textbox(value=video),
)
def slice_audio(audio_path):
print('*'*NUMBER)
print("Slicing audio")
folder_vocals = "vocals"
folder_chunck = "chunks"
if not os.path.exists(folder_vocals):
os.makedirs(folder_vocals)
if not os.path.exists(folder_chunck):
os.makedirs(folder_chunck)
python_file = "slice_audio.py"
command = f"python {python_file} {audio_path} {SECONDS}"
os.system(command)
return (
gr.Textbox(value="Ok")
)
def trascribe_audio(source_languaje, number_of_speakers):
print('*'*NUMBER)
print("Transcript slices")
folder_chunks = "chunks"
python_file = "transcribe.py"
chunks_file = "chunks/output_files.txt"
command = f"python {python_file} {chunks_file} {source_languaje} {number_of_speakers} {DEVICE}"
os.system(command)
with open(chunks_file, 'r') as f:
files = f.read().splitlines()
for file in files:
audios_extension = "mp3"
file_name, _ = file.split(".")
_, file_name = file_name.split("/")
vocal = f'{folder_chunks}/{file_name}.{audios_extension}'
command = f"rm {vocal}"
os.system(command)
return (
gr.Textbox(value="Ok")
)
def concatenate_transcriptions():
print('*'*NUMBER)
print("Concatenate transcriptions")
folder_concatenated = "concatenated_transcriptions"
if not os.path.exists(folder_concatenated):
os.makedirs(folder_concatenated)
chunck_file = "chunks/output_files.txt"
python_file = "concat_transcriptions.py"
command = f"python {python_file} {chunck_file} {SECONDS}"
os.system(command)
with open(chunck_file, 'r') as f:
files = f.read().splitlines()
for file in files:
file_name, _ = file.split(".")
_, file_name = file_name.split("/")
transcriptions_folder = "transcriptions"
transcription_extension = "srt"
command = f"rm {transcriptions_folder}/{file_name}.{transcription_extension}"
os.system(command)
audio_transcribed = "concatenated_transcriptions/download_audio.srt"
return (
gr.Textbox(value="Ok"),
gr.Textbox(value=audio_transcribed),
)
def translate_transcription(original_audio_transcribed_path, source_languaje, target_languaje):
print('*'*NUMBER)
print("Translate transcription")
folder_translated_transcriptions = "translated_transcriptions"
if not os.path.exists(folder_translated_transcriptions):
os.makedirs(folder_translated_transcriptions)
python_file = "translate_transcriptions.py"
command = f"python {python_file} {original_audio_transcribed_path} --source_languaje {source_languaje} --target_languaje {target_languaje} --device {DEVICE}"
os.system(command)
translated_transcription = f"translated_transcriptions/download_audio_{target_languaje}.srt"
transcription_file = "concatenated_transcriptions/download_audio.srt"
if os.path.exists(transcription_file):
command = f"rm {transcription_file}"
os.system(command)
return (
gr.Textbox(value="Ok"),
gr.Textbox(value=translated_transcription)
)
def add_translated_subtitles_to_video(original_video_path, original_audio_path, original_audio_translated_path):
print('*'*NUMBER)
print("Add subtitles to video")
python_file = "add_subtitles_to_video.py"
command = f"python {python_file} {original_audio_translated_path} {original_video_path} {original_audio_path}"
os.system(command)
if os.path.exists(original_video_path):
command = f"rm {original_video_path}"
os.system(command)
if os.path.exists(original_audio_path):
command = f"rm {original_audio_path}"
os.system(command)
if os.path.exists(original_audio_translated_path):
command = f"rm {original_audio_translated_path}"
os.system(command)
if os.path.exists("chunks/output_files.txt"):
command = f"rm chunks/output_files.txt"
os.system(command)
subtitled_video = "videos/download_video_with_subtitles.mp4"
visible = False
return (
gr.Video(value=subtitled_video, visible=True),
gr.Textbox(value="Ok", visible=visible),
gr.Textbox(value="Ok"),
)
def hide_textbobes_progress_info():
visible = False
return (
gr.Textbox(value="Waiting", visible=visible),
gr.Textbox(value="Waiting", visible=visible),
gr.Textbox(value="Waiting", visible=visible),
gr.Textbox(value="Waiting", visible=visible),
gr.Textbox(value="Waiting", visible=visible),
gr.Textbox(value="Waiting", visible=visible),
)
def subtify():
with gr.Blocks(
theme=gr.themes.Default().set
(
body_background_fill=BACKGROUND_COLOR,
body_background_fill_dark=BACKGROUND_COLOR,
body_text_color=PRIMARY_TEXT_COLOR,
body_text_color_dark=PRIMARY_TEXT_COLOR,
body_text_color_subdued=SUBDUED_TEXT_COLOR,
body_text_color_subdued_dark=SUBDUED_TEXT_COLOR,
background_fill_primary=BACKGROUND_PRIMARY_COLOR,
background_fill_primary_dark=BACKGROUND_PRIMARY_COLOR,
background_fill_secondary=BACKGROUND_SECONDARY_COLOR,
background_fill_secondary_dark=BACKGROUND_SECONDARY_COLOR,
border_color_primary=PRIMARY_BODER_COLOR,
border_color_primary_dark=PRIMARY_BODER_COLOR,
block_background_fill=BACKGROUND_PRIMARY_COLOR,
block_background_fill_dark=BACKGROUND_PRIMARY_COLOR,
block_title_text_color=BLOCK_TITLE_TEXT_COLOR,
block_title_text_color_dark=BLOCK_TITLE_TEXT_COLOR,
input_background_fill=INPUT_BACKGROUND_COLOR,
input_background_fill_dark=INPUT_BACKGROUND_COLOR,
input_border_color=INPUT_BORDER_COLOR,
input_border_color_dark=INPUT_BORDER_COLOR,
input_placeholder_color=INPUT_PLACEHOLDER_COLOR,
input_placeholder_color_dark=INPUT_PLACEHOLDER_COLOR,
error_background_fill=ERROR_BACKGROUND_COLOR,
error_background_fill_dark=ERROR_BACKGROUND_COLOR,
error_text_color=ERROR_TEXT_COLOR,
error_text_color_dark=ERROR_TEXT_COLOR,
error_border_color=ERROR_BORDER_COLOR,
error_border_color_dark=ERROR_BORDER_COLOR,
button_secondary_background_fill=BUTTON_SECONDARY_BACKGROUND_COLOR,
button_secondary_background_fill_dark=BUTTON_SECONDARY_BACKGROUND_COLOR,
button_secondary_border_color=BUTTON_SECONDARY_BORDER_COLOR,
button_primary_background_fill_dark=BUTTON_SECONDARY_BORDER_COLOR,
button_secondary_text_color=BUTTON_SECONDARY_TEXT_COLOR,
button_secondary_text_color_dark=BUTTON_SECONDARY_TEXT_COLOR,
)
) as demo:
num_speaker = []
for i in range(100, 0, -1):
num_speaker.append(i)
# Layout
gr.HTML(html_social_media)
gr.HTML("Subtify
")
gr.HTML(html_subtify_logo)
with gr.Row(variant="panel"):
url_textbox = gr.Textbox(placeholder="Add video URL here and wait a moment", label="Video URL", elem_id="video_url", scale=1, interactive=True)
# paste_button = gr.Button(size="sm", icon="icons/paste.svg", value="paste", min_width="10px", scale=0)
delete_button = gr.Button(size="sm", icon="icons/delete.svg", value="clear", min_width="10px", scale=0)
visible = False
auxiliar_block1 = gr.Textbox(label="Auxiliar block 1", elem_id="auxiliar_block1", interactive=False, visible=visible)
with gr.Row(equal_height=False):
image = gr.Image(visible=visible, scale=1)
with gr.Column():
with gr.Row():
source_languaje = gr.Dropdown(visible=visible, label="Source languaje", show_label=True, value="English", choices=language_dict, scale=1, interactive=True, info="Language of the video")
target_languaje = gr.Dropdown(visible=visible, label="Target languaje", show_label=True, value="Español", choices=language_dict, scale=1, interactive=True, info="Language to translate the subtitles")
with gr.Accordion("Advanced settings", open=False, visible=visible) as Advanced_setings:
number_of_speakers = gr.Dropdown(visible=visible, label="Number of speakers", show_label=True, value=10, choices=num_speaker, scale=1, interactive=True, info="Number of speakers in the video, if you don't know, select 10")
subtify_button = gr.Button(size="lg", value="subtify", min_width="10px", scale=0, visible=visible)
auxiliar_block2 = gr.Textbox(placeholder="Waiting", label="Auxiliar block 2", elem_id="auxiliar_block2", interactive=False, visible=visible)
with gr.Row():
video_donwloaded_progress_info = gr.Textbox(placeholder="Waiting", label="Video download progress info", elem_id="video_donwloaded_progress_info", interactive=False, visible=visible)
video_sliced_progress_info = gr.Textbox(placeholder="Waiting", label="Video slice progress info", elem_id="video_sliced_progress_info", interactive=False, visible=visible)
video_transcribed_progress_info = gr.Textbox(placeholder="Waiting", label="Transcribe progress info", elem_id="video_transcribed_progress_info", interactive=False, visible=visible)
transcriptions_concatenated_progress_info = gr.Textbox(placeholder="Waiting", label="Concatenate progress info", elem_id="transcriptions_concatenated_progress_info", interactive=False, visible=visible)
video_translated_progress_info = gr.Textbox(placeholder="Waiting", label="Translate progress info", elem_id="transcription_translated_progress_info", interactive=False, visible=visible)
video_subtitled_progress_info = gr.Textbox(placeholder="Waiting", label="Video subtitle progress info", elem_id="video_subtitled_progress_info", interactive=False, visible=visible)
original_audio_path = gr.Textbox(label="Original audio path", elem_id="original_audio_path", visible=visible)
original_video_path = gr.Textbox(label="Original video path", elem_id="original_video_path", visible=visible)
original_audio_transcribed_path = gr.Textbox(label="Original audio transcribed", elem_id="original_audio_transcribed", visible=visible)
original_audio_translated_path = gr.Textbox(label="Original audio translated", elem_id="original_audio_translated", visible=visible)
subtitled_video = gr.Video(label="Subtitled video", elem_id="subtitled_video", visible=visible, interactive=visible)
auxiliar_block3 = gr.Textbox(placeholder="Waiting", label="Auxiliar block 3", elem_id="auxiliar_block3", interactive=False, visible=visible)
# Events
# paste_button.click(fn=paste_url_from_clipboard, outputs=url_textbox)
delete_button.click(
fn=reset_frontend,
outputs=[
url_textbox,
image,
source_languaje,
target_languaje,
Advanced_setings,
number_of_speakers,
subtify_button,
auxiliar_block2,
video_donwloaded_progress_info,
video_sliced_progress_info,
video_transcribed_progress_info,
transcriptions_concatenated_progress_info,
video_translated_progress_info,
video_subtitled_progress_info,
subtitled_video,
]
)
url_textbox.change(
fn=show_auxiliar_block1,
outputs=[auxiliar_block1]
)
auxiliar_block1.change(
fn=is_valid_url,
inputs=url_textbox,
outputs=[image, source_languaje, target_languaje, Advanced_setings, number_of_speakers, subtify_button]
)
subtify_button.click(
fn=change_visibility_texboxes,
outputs=[auxiliar_block2, video_donwloaded_progress_info, video_sliced_progress_info, video_transcribed_progress_info, transcriptions_concatenated_progress_info, video_translated_progress_info, video_subtitled_progress_info, auxiliar_block1]
)
auxiliar_block2.change(
fn=get_audio_and_video_from_video,
inputs=[url_textbox],
outputs=[video_donwloaded_progress_info, original_audio_path, original_video_path]
)
video_donwloaded_progress_info.change(
fn=slice_audio,
inputs=[original_audio_path],
outputs=[video_sliced_progress_info]
)
video_sliced_progress_info.change(
fn=trascribe_audio,
inputs=[source_languaje, number_of_speakers],
outputs=[video_transcribed_progress_info]
)
video_transcribed_progress_info.change(
fn=concatenate_transcriptions,
outputs=[transcriptions_concatenated_progress_info, original_audio_transcribed_path]
)
transcriptions_concatenated_progress_info.change(
fn=translate_transcription,
inputs=[original_audio_transcribed_path, source_languaje, target_languaje],
outputs=[video_translated_progress_info, original_audio_translated_path]
)
video_translated_progress_info.change(
fn=add_translated_subtitles_to_video,
inputs=[original_video_path, original_audio_path, original_audio_translated_path],
outputs=[subtitled_video, video_subtitled_progress_info, auxiliar_block3]
)
auxiliar_block3.change(
fn=hide_textbobes_progress_info,
outputs=[video_donwloaded_progress_info, video_sliced_progress_info, video_transcribed_progress_info, transcriptions_concatenated_progress_info, video_translated_progress_info, video_subtitled_progress_info]
)
gr.HTML(html_buy_me_a_coffe)
demo.launch()
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--no_ui", action="store_true")
parser.add_argument("--remove_all_files", action="store_true")
args = parser.parse_args()
if args.no_ui:
subtify_no_ui()
elif args.remove_all_files:
remove_all_files()
else:
subtify()
