File size: 4,222 Bytes
d5ee97c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 |
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import random\n",
"import shutil\n",
"import sys\n",
"\n",
"libri_path = \"....../LibriTTS\" # absolute path to TensorFlowTTS.\n",
"dataset_path = \"....../libritts\" # Change to your paths. This is a output of re-format dataset.\n",
"subset = \"train-clean-100\""
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"with open(os.path.join(libri_path, \"SPEAKERS.txt\")) as f:\n",
" data = f.readlines()\n",
" \n",
"dataset_info = {}\n",
"max_speakers = 20 # Max number of speakers to train on\n",
"min_len = 20 # Min len of speaker narration time\n",
"max_file_len = 11 # max audio file lenght\n",
"min_file_len = 2 # min audio file lenght"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"possible_dataset = [i.split(\"|\") for i in data[12:] if i.split(\"|\")[2].strip() == subset and float(i.split(\"|\")[3].strip()) >= min_len]"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"ids = [i[0].strip() for i in possible_dataset]"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"import soundfile as sf"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"possible_map = {}\n",
"subset_path = os.path.join(libri_path, subset)\n",
"for i in os.listdir(subset_path):\n",
" if i in ids:\n",
" id_path = os.path.join(subset_path, i)\n",
" id_dur = 0\n",
" id_included = []\n",
" \n",
" for k in os.listdir(id_path):\n",
" for j in os.listdir(os.path.join(id_path, k)):\n",
" if \".wav\" in j:\n",
" f_path = os.path.join(id_path, k, j)\n",
" sf_file = sf.SoundFile(f_path)\n",
" dur = len(sf_file) / sf_file.samplerate\n",
" if max_file_len < dur or dur < min_file_len:\n",
" continue\n",
" else:\n",
" id_included.append(f_path)\n",
" id_dur += dur\n",
" possible_map[i] = {\"dur\": id_dur, \"included\": id_included}\n"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"poss_speakers = {k: v[\"included\"] for k, v in possible_map.items() if v[\"dur\"]/60 >= min_len}"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"to_move = list(poss_speakers.keys())\n",
"random.shuffle(to_move)\n",
"to_move = to_move[:max_speakers]"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"for sp_id, v in poss_speakers.items():\n",
" if sp_id in to_move:\n",
" for j in v:\n",
" f_name = j.split(os.path.sep)[-1]\n",
" text_f_name = f_name.split(\".wav\")[0] + \".txt\"\n",
" os.makedirs(os.path.join(dataset_path, sp_id), exist_ok=True)\n",
" shutil.copy(j, os.path.join(dataset_path, sp_id, f_name))\n",
" shutil.copy(j.replace(\".wav\", \".normalized.txt\"), os.path.join(dataset_path, sp_id, text_f_name))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"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": "3.7.7"
}
},
"nbformat": 4,
"nbformat_minor": 4
}
|