SARdBScene

Signal Analysis Research group's database of Source Counting and audio ENvironment Evaluation (SARdBScene)

If using this dataset please cite the following paper: 

M. Nigro and S. Krishnan, “SARdBScene: Dataset and ResNet Baseline for Audio Scene Source Counting and Analysis,” in ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2023

Please see here for more details including baseline code and auxiliary files

Audio scene analysis dataset mainly focused on audio source counting among other audio-based tasks.

'SARdBScene' folder contains audio mixtures, 'SARdBScene_annotations' contains labels for each mixture concerning source counting and other tasks; source audio used in the synthesizing process are also listed in here for cross-referencing. 'transcripts' contains .txt files corresponding to the speech transcripts for each mixture; there are more transcripts than audio files, and transcripts corresponding to an audio mixture should be cross referenced with the appropriate SARdBScene_annotations folder files