JMIR Publications

ABSTRACT

Background:

Qualitative self- or parent-reports used in assessing children’s behavioral disorders are often inconvenient to collect and can be misleading due to missing information. A data-driven approach to quantify behavioral disorder could alleviate these concerns. This study proposes a machine learning approach to identify screams in voice recordings that avoids the need to gather large amounts of clinical data for model training.

Objective:

The goal of this study is to evaluate if a machine learning model trained only on publicly available audio datasets can be used to detect screaming sounds in audio streams captured in an at-home setting.

Methods:

Two sets of audio samples were prepared to evaluate the model: a subset of the publicly available AudioSet dataset, and a set of audio data extracted from the TV show Supernanny, which was chosen for its similarity to clinical data. Scream events were manually annotated for the Supernanny data and existing annotations were refined for the AudioSet data. Audio feature extraction was performed with a convolutional neural network pre-trained on AudioSet. A gradient-boosted tree model was trained and cross-validated for scream classification on the AudioSet data and then validated independently on the Supernanny audio.

Results:

On the held-out AudioSet clips, the model achieved an ROC-AUC of 0.86. The same model applied to three full episodes of Supernanny audio achieves an ROC-AUC of 0.95 and an average precision (positive predictive value) of 42% despite screams only making up 1.3% of the total runtime.

Conclusions:

These results suggest that a scream-detection model trained with publicly available data could be valuable for monitoring clinical recordings and identifying tantrums, as opposed to depending on collecting costly privacy-protected clinical data for model training.