Upload 2 files

download_datasets_in_wav_or_mp3_and_create_csv.ipynb

@@ -0,0 +1,67 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset\n",
+    "import soundfile as sf, os, pandas as pd, re\n",
+    "from tqdm import tqdm\n",
+    "\n",
+    "dataset = load_dataset(\"mozilla-foundation/common_voice_17_0\", \"ja\", split=\"other\", trust_remote_code=True, streaming=True, token=\"\")\n",
+    "\n",
+    "name = \"other\"\n",
+    "ouput_dir = \"./datasets/CV17/\"\n",
+    "output_file = 'metadata.csv'\n",
+    "os.makedirs(ouput_dir + name, exist_ok=True)\n",
+    "folder_path = ouput_dir + name # Create a folder to store the audio and transcription files\n",
+    "\n",
+    "char = '[　０１２３４５６７８９ａｂｃｄｅｆｇｈｉｊｋｌｍｎｏｐｑｒｓｔｕｖｗｘｙｚＡＢＣＤＥＦＧＨＩＪＫＬＭＮＯＰＱＲＳＴＵＶＷＸＹＺabcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890]'\n",
+    "special_characters = '[♬「」？！“%‘”~♪…?!゛＃＄％＆（）＊＋：；〈＝〉＠＾＿｛｜｝～\"█♩♫』『.;:<>_()*&^$#@`, ]'\n",
+    "\n",
+    "dsa = (dataset\n",
+    "       .filter(lambda sample: bool(sample[\"sentence\"]))\n",
+    "       .filter(lambda sample: not re.search(char, sample[\"sentence\"]))\n",
+    "       .filter(lambda sample: sample[\"down_votes\"] == 0)\n",
+    "       )\n",
+    "\n",
+    "for i, sample in tqdm(enumerate(dsa)): # Process each sample in the filtered dataset\n",
+    "    audio_sample = name + f'_{i}.wav' # or wav\n",
+    "    audio_path = os.path.join(folder_path, audio_sample)\n",
+    "    transcription_path = os.path.join(folder_path, output_file)  # Path to save transcription file \n",
+    "    if not os.path.exists(audio_path):\n",
+    "        sf.write(audio_path, sample['audio']['array'], sample['audio']['sampling_rate'])\n",
+    "        sample[\"audio_length\"] = len(sample[\"audio\"][\"array\"]) / sample[\"audio\"][\"sampling_rate\"]  # Get audio length, remove if not needed\n",
+    "        with open(transcription_path, 'a', encoding='utf-8') as transcription_file: # Save transcription file    \n",
+    "            transcription_file.write(audio_sample+\",\") # Save transcription file name  \n",
+    "            sample[\"sentence\"] = re.sub(special_characters,'', sample[\"sentence\"])\n",
+    "            transcription_file.write(sample['sentence']) # Save transcription \n",
+    "            transcription_file.write(str(\",\"+str(sample['audio_length']))) # Save audio length, remove if not needed\n",
+    "            transcription_file.write('\\n') "
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

extract_audio_sentences_from_mp4_line_by_line_and_create_csv.ipynb

@@ -0,0 +1,330 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import os, io, re, sys, time, datetime, wave, contextlib, librosa\n",
+    "from glob import glob\n",
+    "import numpy as np\n",
+    "from moviepy.editor import *\n",
+    "import soundfile as sf\n",
+    "from pydub import AudioSegment\n",
+    "\n",
+    "def create_directories():\n",
+    "    slice_path = './ready_for_slice'\n",
+    "    if not os.path.exists(slice_path):\n",
+    "        try:\n",
+    "            os.mkdir(slice_path)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %slice_path)\n",
+    "    sliced_audio = './sliced_audio'\n",
+    "    if not os.path.exists(sliced_audio):\n",
+    "        try:\n",
+    "            os.mkdir(sliced_audio)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %sliced_audio)\n",
+    "\n",
+    "    merged_csv_files = './merged_csv'\n",
+    "    if not os.path.exists(merged_csv_files):\n",
+    "        try:\n",
+    "            os.mkdir(merged_csv_files)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %merged_csv_files)\n",
+    "\n",
+    "    final_csv_files = './final_csv'\n",
+    "    if not os.path.exists(final_csv_files):\n",
+    "        try:\n",
+    "            os.mkdir(final_csv_files)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %final_csv_files)\n",
+    "            \n",
+    "    audio = './audio'\n",
+    "    if not os.path.exists(audio):\n",
+    "        try:\n",
+    "            os.mkdir(audio)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %audio)\n",
+    "            \n",
+    "    srt_files = './srt_files'\n",
+    "    if not os.path.exists(srt_files):\n",
+    "        try:\n",
+    "            os.mkdir(srt_files)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %srt_files)\n",
+    "\n",
+    "def merge_csv(path):\n",
+    "    print('Merging csv-files with transcriptions')\n",
+    "    csv_combined = pd.DataFrame()\n",
+    "    for entry in glob (path+'*.csv'):\n",
+    "        df = pd.read_csv(entry)\n",
+    "        csv_combined = csv_combined.append(df)\n",
+    "\n",
+    "    csv_combined.to_csv('./merged_csv/Full_Transcript.csv', header=True, index=False, encoding='utf-8')\n",
+    "    print('All csv-files merged')\n",
+    "\n",
+    "def change_encoding(srt):\n",
+    "    with io.open(srt, 'r', encoding='utf-8') as f:\n",
+    "        text = f.read()\n",
+    "        # process Unicode text\n",
+    "    with io.open(srt, 'w', encoding='utf-8') as f:\n",
+    "        f.write(text)\n",
+    "\n",
+    "def convert_srt_to_csv(file):\n",
+    "    with open(file, 'r', encoding='utf-8') as h:\n",
+    "        sub = h.readlines()   #returns list of all lines\n",
+    "\n",
+    "    re_pattern = r'[0-9]{2}:[0-9]{2}:[0-9]{2},[0-9]{3} --> [0-9]{2}:[0-9]{2}:[0-9]{2},[0-9]{3}'\n",
+    "    regex = re.compile(re_pattern)\n",
+    "    # Get start times\n",
+    "    times = list(filter(regex.search, sub))\n",
+    "    end_times = [time.split('--> ')[1] for time in times] #returns a list\n",
+    "    start_times = [time.split(' ')[0] for time in times]  #returns a list\n",
+    "\n",
+    "    # Get lines\n",
+    "    lines = [[]]\n",
+    "    for sentence in sub:\n",
+    "        if re.match(re_pattern, sentence):\n",
+    "            lines[-1].pop()\n",
+    "            lines.append([])\n",
+    "        else:\n",
+    "            lines[-1].append(sentence)\n",
+    "\n",
+    "    lines = lines[1:]   #all text in lists\n",
+    "\n",
+    "    column_names = ['id','start_times', 'end_times', 'sentence']\n",
+    "    df_text = pd.DataFrame(columns=column_names)\n",
+    "\n",
+    "    df_text['start_times'] = start_times\n",
+    "    df_text['end_times'] = end_times\n",
+    "    df_text['sentence'] = [\" \".join(i).replace('\\n', '') for i in lines]\n",
+    "    df_text['end_times'] = df_text['end_times'].replace(r'\\n', '', regex=True)\n",
+    "\n",
+    "    df_text['id'] = np.arange(len(df_text))\n",
+    "    id_extension = os.path.basename(file).replace('.srt', '_')\n",
+    "    id_extension = id_extension.replace(' ', '_')\n",
+    "    id_extension = id_extension.replace('-', '_')\n",
+    "    id_extension = id_extension.replace('.', '_')\n",
+    "    id_extension = id_extension.replace('__', '_')\n",
+    "    id_extension = id_extension.replace('___', '_')\n",
+    "    df_text['id'] = id_extension +  df_text['id'].map(str)\n",
+    "    file_extension = id_extension[:-1]\n",
+    "\n",
+    "    def convert_to_ms(time):\n",
+    "        h_ms = int(time[:2])*3600000\n",
+    "        m_ms = int(time[3:5])*60000\n",
+    "        s_ms = int(time[6:8])*1000\n",
+    "        ms = int(time[9:12])\n",
+    "        ms_total = h_ms + m_ms + s_ms + ms\n",
+    "        return(ms_total)\n",
+    "\n",
+    "    def conv_int(start):\n",
+    "        new_start = int(start)\n",
+    "        return(new_start)\n",
+    "\n",
+    "    df_text['start_times'] = df_text['start_times'].apply(convert_to_ms)\n",
+    "    df_text['end_times'] = df_text['end_times'].apply(convert_to_ms)\n",
+    "    df_text['start_times'] = df_text['start_times'].apply(conv_int)\n",
+    "    df_text.to_csv('./ready_for_slice/' + file_extension + '.csv', index=False, header=True, encoding='utf-8-sig')\n",
+    "\n",
+    "def wmv_to_wav(entry):\n",
+    "    video = VideoFileClip(entry)\n",
+    "    audio = video.audio\n",
+    "    filename = os.path.basename(entry)\n",
+    "    filename = filename.replace(' ', '_')\n",
+    "    filename = filename.replace('-', '_')\n",
+    "    filename = filename.replace('.', '_')\n",
+    "    filename = filename.replace('__', '_')\n",
+    "    filename = filename.replace('___', '_')\n",
+    "    filename = filename[:-4] + '.wav'\n",
+    "    audio.write_audiofile('./audio/' +filename)\n",
+    "\n",
+    "def mp4_to_wav(entry):\n",
+    "    video = VideoFileClip(entry)\n",
+    "    #extract audio from video\n",
+    "    audio = video.audio\n",
+    "    filename = os.path.basename(entry)\n",
+    "    filename = filename.replace(' ', '_')\n",
+    "    filename = filename.replace('-', '_')\n",
+    "    filename = filename.replace('.', '_')\n",
+    "    filename = filename.replace('__', '_')\n",
+    "    filename = filename.replace('___', '_')\n",
+    "    filename = filename[:-4] + '.wav'\n",
+    "    #filename = filename[:-4]+'.wav'\n",
+    "    #filename = filename[:10] + '_' + filename[-9:]\n",
+    "    audio.write_audiofile('./audio/' +filename)\n",
+    "\n",
+    "def pre_process_audio(audio_path):\n",
+    "    path_audio_processed = './ready_for_slice/'\n",
+    "    if not os.path.exists(path_audio_processed):\n",
+    "        try:\n",
+    "            os.mkdir(path_audio_processed)\n",
+    "        except OSError:\n",
+    "            print('Creation of directory %s failed' %path_audio_processed)\n",
+    "        else:\n",
+    "            print('Successfully created the directory %s' %path_audio_processed)\n",
+    "    start_sub = time.time()\n",
+    "    n = 0\n",
+    "    print('Downsampling wav files...')\n",
+    "    for file in os.listdir(audio_path):\n",
+    "        if(file.endswith('.wav')):\n",
+    "            try:\n",
+    "                nameSolo_1 = file.rsplit('.', 1)[0]\n",
+    "                y, s = librosa.load(audio_path + file, sr=16000) # Downsample 44.1kHz to 8kHz\n",
+    "                sf.write(path_audio_processed + nameSolo_1 + '.wav', y, s)\n",
+    "                n = n+1\n",
+    "                print('File ', n , ' completed:', nameSolo_1)\n",
+    "            except EOFError as error:\n",
+    "                next\n",
+    "\n",
+    "    s = 0\n",
+    "    print('Changing bit pro sample...')\n",
+    "    for file in os.listdir(path_audio_processed):\n",
+    "        if(file.endswith('.wav')):\n",
+    "            try:\n",
+    "                nameSolo_2 = file.rsplit('.', 1)[0]\n",
+    "                #nameSolo_2 = nameSolo_2.replace('')\n",
+    "                data, samplerate = sf.read(path_audio_processed + file)\n",
+    "                sf.write(path_audio_processed + nameSolo_2 + '.wav', data, samplerate, subtype='PCM_16')\n",
+    "                s = s + 1\n",
+    "                print('File ' , s , ' completed')\n",
+    "            except EOFError as error:\n",
+    "                next\n",
+    "\n",
+    "    end_sub = time.time()\n",
+    "    print('The script took ', end_sub-start_sub, ' seconds to run')\n",
+    "    \n",
+    "def create_DS_csv (path):\n",
+    "    print('Extracting filepath and -size for every .wav file in ./sliced_audio')\n",
+    "    data = pd.DataFrame(columns=['file_name', 'duration'])\n",
+    "    df = pd.DataFrame(columns=['file_name', 'duration'])\n",
+    "\n",
+    "    for entry in glob(path +'*.wav'):\n",
+    "        filename = os.path.basename(entry)\n",
+    "        with contextlib.closing(wave.open(entry, 'rb')) as f:\n",
+    "            frames = f.getnframes()\n",
+    "            rate = f.getframerate()\n",
+    "            duration = frames / float(rate)\n",
+    "        df['file_name'] = [filename]\n",
+    "        df['duration'] = [duration]\n",
+    "        data = data.append(df)\n",
+    "    data.to_csv('./merged_csv/Filepath_Filesize.csv', header=True, index=False, encoding='utf-8')\n",
+    "\n",
+    "def split_files(item, wav_item):\n",
+    "    song = AudioSegment.from_wav(wav_item)\n",
+    "    df = pd.read_csv(item)\n",
+    "\n",
+    "    def audio_split(df):\n",
+    "        split = song[df['start_times']:df['end_times']]\n",
+    "        split.export('./sliced_audio/' + df['id'] + '.wav', format ='wav')\n",
+    "\n",
+    "    df.apply(audio_split, axis=1)\n",
+    "\n",
+    "def merge_transcripts_and_wav_files(transcript_path, DS_csv):\n",
+    "    df_final = pd.DataFrame()\n",
+    "    df_transcripts = pd.read_csv(transcript_path)\n",
+    "    df_files = pd.read_csv(DS_csv)\n",
+    "    def remove_path(path):\n",
+    "        path = path.split('/')[-1]\n",
+    "        return path\n",
+    "    df_files['id'] = df_files['file_name'].apply(remove_path)\n",
+    "    #filter out duration of less than 10 seconds\n",
+    "    def convert(duration):\n",
+    "        time = float(duration)\n",
+    "        return time\n",
+    "    df_files['duration'] = df_files['duration'].apply(convert)\n",
+    "    df_files = df_files[df_files['duration']<10.00]\n",
+    "    #drop unnecessary columns\n",
+    "    df_transcripts.drop(['start_times','end_times'], axis=1, inplace=True)\n",
+    "    df_files.drop(['duration'], axis=1, inplace=True)\n",
+    "    df_files['id'] = df_files['id'].replace('.wav', '', regex=True)\n",
+    "    #merge on column id\n",
+    "    df_final = pd.merge(df_transcripts, df_files, on='id')\n",
+    "    df_final.drop(['id'], axis=1, inplace=True)\n",
+    "    #rearrange columns\n",
+    "    df_final = df_final[['file_name', 'sentence']]\n",
+    "    df_final.to_csv('./final_csv/metadata.csv', header=True, index=False, encoding='utf-8')\n",
+    "    \n",
+    "    create_directories()\n",
+    "    \n",
+    "print(\"Put your video or audio files into the audio folder and srt files into the srt_files folder when you're ready...\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "start_time = time.time()\n",
+    "srt_path = './srt_files'\n",
+    "audio_path = './audio/'\n",
+    "srt_counter = len(glob('./srt_files/' + '*.srt'))\n",
+    "\n",
+    "#Extracting information from srt-files to csv\n",
+    "print('Extracting information from srt_file(s) to csv_files')\n",
+    "for file in glob('./srt_files/*.srt'):\n",
+    "    convert_srt_to_csv(file)\n",
+    "print('%s-file(s) converted and saved as csv-files to ./csv' %srt_counter)\n",
+    "print('---------------------------------------------------------------------')\n",
+    "\n",
+    "#extract audio (wav) from mp4\n",
+    "for entry in glob('./audio/*.mp4'):\n",
+    "    mp4_to_wav(entry)\n",
+    "print('MP4 to WAV convert complete')\n",
+    "print('---------------------------------------------------------------------')\n",
+    "\n",
+    "\n",
+    "#Pre-process audio for folder in which wav files are stored\n",
+    "pre_process_audio(audio_path)\n",
+    "print('Pre-processing of audio files is complete.')\n",
+    "print('---------------------------------------------------------------------')\n",
+    "\n",
+    "print('Slicing audio according to start- and end_times of transcript_csvs...')\n",
+    "for item in glob('./ready_for_slice/*.csv'):\n",
+    "    wav_item = item.replace('.csv','.wav')\n",
+    "    if os.path.exists(wav_item):\n",
+    "        split_files(item, wav_item)\n",
+    "    else:\n",
+    "        next\n",
+    "wav_counter = len(glob('./sliced_audio/' + '*.wav'))\n",
+    "print('Slicing complete. {} files in dir \"sliced_audio\"'.format(wav_counter))\n",
+    "print('---------------------------------------------------------------------')\n",
+    "\n",
+    "create_DS_csv('./sliced_audio/')\n",
+    "\n",
+    "#now join all seperate csv files\n",
+    "merge_csv('./ready_for_slice/')\n",
+    "print('Merged csv with all transcriptions created.')\n",
+    "print('---------------------------------------------------------------------')\n",
+    "transcript_path = './merged_csv/Full_Transcript.csv'\n",
+    "DS_path = './merged_csv/Filepath_Filesize.csv'\n",
+    "merge_transcripts_and_wav_files(transcript_path, DS_path)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}