Sneeze Counter

Source acknowledgments

Sound model
& credits.

The on-device sound classifier was developed using licensed recordings from VocalSound and Freesound contributors. No training recordings are included in the app or this download.

Model and reproducible recipe

Sun Ryse Software LLC provides the selected model and its training recipe under Creative Commons Attribution-ShareAlike 4.0. The separate download has no account requirement or digital restrictions. Apple’s frameworks and the app’s other code retain their own terms.

Download model & recipe

Read the model attribution and license notice
Read the individual recording credits

VocalSound

Yuan Gong, Jin Yu, and James Glass. VocalSound: A Dataset for Improving Human Vocal Sounds Recognition. ICASSP 2022, pages 151–155. Research paper · Original dataset. The original dataset’s CC BY-SA 4.0 terms are retained. The training recipe identifies the selected published TRAIN recordings; the published TEST recordings were reserved for evaluation.

Freesound contributors

The individual credits identify the selected authors, source recordings, and CC0 or CC BY licenses. Training preparation converted audio to mono WAV and applied the documented padding. The model is a learned classifier; it is not a collection of the source recordings.

These acknowledgments do not imply endorsement by the researchers, recording contributors, or Apple. Sound detection remains experimental and does not establish who sneezed, why they sneezed, or a medical diagnosis.