{ "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import pandas as pd\n", "from libris2s_dataset_pt import Libris2sDataset" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ " DE_audio \\\n", "book_id \n", "10 DE\\10.buchsederpandora_1501_librivox\\sentence_... \n", "10 DE\\10.buchsederpandora_1501_librivox\\sentence_... \n", "10 DE\\10.buchsederpandora_1501_librivox\\sentence_... \n", "10 DE\\10.buchsederpandora_1501_librivox\\sentence_... \n", "10 DE\\10.buchsederpandora_1501_librivox\\sentence_... \n", "\n", " EN_audio score \\\n", "book_id \n", "10 EN\\10.pandora\\sentence_level_audio\\00001-f0000... 0.508904 \n", "10 EN\\10.pandora\\sentence_level_audio\\00001-f0000... 0.625685 \n", "10 EN\\10.pandora\\sentence_level_audio\\00001-f0000... 0.262500 \n", "10 EN\\10.pandora\\sentence_level_audio\\00001-f0000... 0.500168 \n", "10 EN\\10.pandora\\sentence_level_audio\\00001-f0000... 0.845455 \n", "\n", " DE_transcript \\\n", "book_id \n", "10 Prachtvoller Saal in deutscher Renaissance mit... \n", "10 In der Mitte unter der Galerie befindet sich d... \n", "10 Links vorn ein kleiner Serviertisch, daneben e... \n", "10 Der Saal ist durch eine auf dem Mitteltisch st... \n", "10 Alwa Schön geht vor der Eingangstür auf und ni... \n", "\n", " EN_transcript \n", "book_id \n", "10 ACT I The hall of EARTHSPIRIT, , feebly lighte... \n", "10 ACT I The hall of EARTHSPIRIT, , feebly lighte... \n", "10 The firescreen and the chair by the ottoman ar... \n", "10 Down left is a small teatable, with a coffeepo... \n", "10 In this chair, deep in cushions, with a plaid ... \n", "total number of samples in the alignment file: 25604\n" ] } ], "source": [ "# display the dataframe\n", "allignment = pd.read_csv(\"alignments/all_de_en_alligned_cleaned.csv\",index_col=0)\n", "print(allignment.head())\n", "print(\"total number of samples in the alignment file:\",len(allignment))" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "number of samples loaded in the dataset: 25604\n" ] } ], "source": [ "dataset = Libris2sDataset(\n", " data_dir=\".\",\n", " split=\"alignments/all_de_en_alligned_cleaned.csv\", \n", ")\n", "print(\"number of samples loaded in the dataset:\",len(dataset))" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "# Get a sample\n", "sample = dataset[123]\n", "\n", "# Access the data\n", "de_audio = sample['de_audio'] # German audio waveform\n", "en_audio = sample['en_audio'] # English audio waveform\n", "de_sample_rate = sample['de_sample_rate']\n", "en_sample_rate = sample['en_sample_rate']\n", "de_text = sample['de_transcript'] # German transcript\n", "en_text = sample['en_transcript'] # English transcript\n", "score = sample['alignment_score'] # Alignment quality score" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "from IPython.display import display, HTML, Audio\n", "\n", "def display_alignment_sample(sample):\n", " de_audio = sample['de_audio']\n", " en_audio = sample['en_audio']\n", " de_sample_rate = sample['de_sample_rate']\n", " en_sample_rate = sample['en_sample_rate']\n", " de_text = sample['de_transcript']\n", " en_text = sample['en_transcript']\n", " score = sample['alignment_score']\n", "\n", " html = f\"\"\"\n", "

🔗 Alignment Sample

\n", "

Alignment Score: {score:.3f}

\n", " \n", "

German

\n", "

Transcript: {de_text}

\n", " \"\"\"\n", " display(HTML(html))\n", " display(Audio(de_audio.numpy(), rate=de_sample_rate))\n", " \n", " html = f\"\"\"\n", "

English

\n", "

Transcript: {en_text}

\n", " \"\"\"\n", " display(HTML(html))\n", " display(Audio(en_audio.numpy(), rate=en_sample_rate))\n" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "

🔗 Alignment Sample

\n", "

Alignment Score: 0.907

\n", "\n", "

German

\n", "

Transcript: Wenn ich eben Fräulein von Geschwitz darum bat, meine Hilfe anzunehmen, so geschah es gewiß nicht, um Ihre unersättliche Goldgier aufzustacheln.

\n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", " \n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "

English

\n", "

Transcript: When I asked Fräulein von Geschwitz just now to accept my help, it certainly was not to incite your insatiable avarice.

\n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", " \n", " " ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "display_alignment_sample(sample)" ] } ], "metadata": { "kernelspec": { "display_name": "cerba", "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": "3.12.6" } }, "nbformat": 4, "nbformat_minor": 2 }