IWSLT 2022-2023 Shared Task Training, Development and Test Set

<h3>Introduction</h3>
<p>IWSLT 2022 - 2023 Shared Task Training, Development and Test Set (LDC2025S05) was developed by the Linguistic Data Consortium (LDC). It contains 210 hours of Tunisian Arabic conversational telephone speech, transcripts and their English translations covering 175 hours of that speech, speaker metadata, and documentation. This material constitutes the training, development and test data used in the International Conference on Spoken Language Translation (IWSLT) <a href="https://iwslt.org/2022/dialect">Dialectal Speech Translation task (2022)</a> and the <a href="https://iwslt.org/2023/low-resource">Dialectal and Low-resource track (2023)</a>.</p>
<h3>Data</h3>
<p>The telephone speech was collected by LDC in 2016-2017 from native speakers of Tunisian Arabic in Tunis. Speakers were recruited to make telephone calls to people in their social networks from a variety of noise conditions and handsets. The calls were recorded using a robot operator system that captured digital audio samples directly from the regional public telephone network with the informed consent of participants. The audio files are two-channel recordings across 1,188 conversations.</p>
<p>Transcripts are orthographic following <a href="../../../LDC2004L02">Buckwalter</a> transliteration. IPA (International Phonetic Alphabet) transcripts were added to a subset of the data. All transcribed segments were translated into English. Further information on the transcription and translation methodologies is contained in the documentation accompanying this release.</p>
<p>Speech data is presented as FLAC-compressed MS-WAV files in 16-bit 8 kHz PCM format. All text data is UTF-8 encoded.</p>
<h3>Updates</h3>
<p>No updates at this time.</p>