41 datasets found

A dataset of audio recordings used for training and evaluation of the Binaural Angular Separation Network.

A combination of LibriVox and Freesound datasets used for training and evaluation.

The Voxceleb2 dataset is a large-scale speaker recognition dataset, containing 2442 hours raw speech from 6112 speakers.

Polyphonic music tasks characterized by multivariate time-series. The analysis is performed in terms of efficiency and prediction accuracy on 4 polyphonic music tasks.

Automated Audio captioning (AAC) is a cross-modal translation task that aims to use natural language to describe the content of an audio clip.

The dataset used for training the model consists of 13 pieces of Mozart, 989 pieces for validation, and 11,821 pieces for testing.

The TSP speech database is a dataset of speech recordings.

The dataset used in this paper is the Isolet dataset, which contains 4,000 13-channel audio recordings of 100 speakers.

The PANNS dataset is a large-scale audio classification dataset.

The Freesound dataset consists of 18,873 audio files, each assigned one of the 41 unique audio events from the Google's Audioset Ontology.

Few-shot learning aims to generalize unseen classes that appear during testing but are unavailable during training. The performance of prototypical networks in extreme few-shot...

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.

The TEDLIUM2 dataset is a large corpus of audio recordings of human speech, with a focus on speech recognition tasks.

The dataset used in the paper is the WSJ0-2mix dataset, which contains 30 hours of training data and 10 hours of validation data generated from the WSJ0 dataset. The speech...

Deep Matrix Approximately Nonlinear Decomposition (DEMAND) employs Adam and an alternative optimization strategy that is well suited to optimize convex/alternative convex...

The WavCaps dataset contains chatGPT-assisted weakly-labeled audio captioning data.

The COVID-19 Identification ResNet (CIdeR) dataset consists of 517 crowdsourced coughing and breathing audio recordings from 355 participants, of which 62 participants had tested...

Speech enhancement dataset

The dataset used in this paper is a collection of phonetically and phonologically local allophonic distribution in English, where voiceless stops surface as aspirated...

The VGGSound dataset is a large-scale audio-visual dataset containing 10,000 10-second video clips with corresponding audio files.