12 datasets found

Groups: Speech Corpus Formats: JSON

Filter Results

  • AISHELL-1

    The AISHELL-1 dataset is a Mandarin speech corpus, consisting of 178 hours of speech, with 11 domains and 400 speakers from different accent areas in China.

  • TED-LIUM 3

    TED-LIUM 3 (TL3) is a TED talks dataset. Speaker adaptation data for TL3 was divided randomly, where 2/5 was divided into the train set, 1/5 was divided into the dev set, and...

  • TIMIT Corpus

    The TIMIT corpus is a large database of speech recordings used for speaker recognition and speech synthesis tasks.

  • TIMIT

    The TIMIT corpus is a widely used benchmark for speech recognition tasks. It contains 3,696 training utterances from 462 speakers, excluding the SA sentences. The core test set...

  • HKUST/MTS: A Very Large Scale Mandarin Telephone Speech Corpus

    The HKUST dataset is a large dataset of speech recordings, each containing a single speaker speaking a sentence.

  • The Wall Street Journal Corpus

    The WSJ dataset is a large dataset of speech recordings, each containing a single speaker speaking a sentence.

  • TIMIT Acoustic-Phonetic Continuous Speech Corpus

    The TIMIT acoustic-phonetic continuous speech corpusCD-ROM contains a large collection of speech samples from 250 male and 250 female speakers.

  • Librispeech

    The Librispeech dataset is a large-scale speaker-dependent speech corpus containing 1080 hours of speech, 5600 utterances, and 1000 speakers.

  • LibriLight

    The dataset used in this paper is a large-scale production ASR system, which includes multi-domain (MD) data sets in English. The MD data sets include medium-form (MF) and...

  • VCTK

    Voice conversion (VC) is a technique that alters the voice of a source speaker to a target style, such as speaker identity, prosody, and emotion, while keeping the linguistic...

  • VoxCeleb1

    Speaker recognition aims to identify speaker information from input speech. A type of speaker recognition is speaker verification (SV). It determines whether the test speaker's...

  • LibriTTS

    A popular text-based VC approach is to use an automatic speech recognition (ASR) model to extract phonetic posteriorgram (PPG) as content representation.