Reproduce “Enhancing Infant Crying Detection with Gradient Boosting for Improved Emotional and Mental Health Diagnostics” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Kyunghun Lee, Lauren M. Henry, Eleanor Hansen, Elizabeth Tandilashvili, Lauren S. Wakschlag, Elizabeth Norton, Daniel S. Pine, Melissa A. Brotman and 1 more

A framework combines deep learning Wav2Vec embeddings with traditional audio features and Gradient Boosting Machines to detect infant crying. This approach achieves a 0.931 AUC on the HomeBank English deBarbaro Cry Corpus, exceeding previous state-of-the-art methods.

Key insights

  • The integration of deep learning Wav2Vec embeddings with traditional acoustic features creates a superior and complementary feature set for infant cry detection.
  • Gradient Boosting Machines provide a robust and high-performing classification algorithm for this complex audio task, outperforming both re-implemented state-of-the-art and SVM baselines.
  • An automated framework for infant cry detection, leveraging combined features and GBM, offers a scalable and objective alternative to manual annotation and subjective reporting for large-scale naturalistic audio datasets.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.