PhilippXXY/AudibleLight_Eigenmike32-5_DCASE-STARSS23_Dataset

AudibleLight Eigenmike32-5 DCASE-STARSS23 Dataset

This dataset contains 121 synthetic spatial audio scenes — 111 for training and 10 for evaluation — of 60 seconds each, generated with the AudibleLight dataset generator (DOI). Each scene is rendered as five independent simulated Eigenmike32 captures, with 32 channels per capture, resulting in 570 minutes of multichannel audio in total at 24 kHz.

Foreground Audio

Foreground events are sampled from ESC-50: Dataset for Environmental Sound Classification for the training and test set, converted to mono and transformed to 24 kHz. For each scene, the foreground sample selection and event duration are randomly chosen.

Background Noise and Mixing

Background noise is added and randomly selected from white, pink, or Gaussian noise. Scene parameters such as Signal-to-Noise Ratio, background noise floor (dB), and applied augmentations vary and are sampled randomly across scenes.

Acoustic Simulation and Environments

The ray-tracing simulation is performed on a randomly selected Gibson mesh per scene. Additional synthesis parameters such as source location and other configuration values are sampled randomly to increase scene diversity.

Ground Truth

Ground truth is stored in DCASE/STARSS23 format. Class labels are set to 0 for all events, since semantic class identity is not required for the downstream applications targeted by this dataset.