Privacy-Enhancing Infant Cry Classification with Federated Transformers and Denoising Regularization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Owino, Geofrey, Shibwabo, Bernard
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912767279104000
author Owino, Geofrey
Shibwabo, Bernard
author_facet Owino, Geofrey
Shibwabo, Bernard
contents Infant cry classification can aid early assessment of infant needs. However, deployment of such solutions is limited by privacy concerns around audio data, sensitivity to background noise, and domain shift across recording environments. We present an end-to-end infant cry analysis pipeline that integrates a denoising autoencoder (DAE), a convolutional tokenizer, and a Transformer encoder trained using communication-efficient federated learning (FL). The system performs on-device denoising, adaptive segmentation, post hoc calibration, and energy-based out-of-distribution (OOD) abstention. Federated training employs a regularized control variate update with 8-bit adapter deltas under secure aggregation. Using the Baby Chillanto and Donate-a-Cry datasets with ESC-50 noise overlays, the model achieves a macro F1 score of 0.938, an AUC of 0.962, and an Expected Calibration Error (ECE) of 0.032, while reducing per-round client upload from approximately 36 to 42 MB to 3.3 MB. Real-time edge inference on an NVIDIA Jetson Nano (4 GB, TensorRT FP16) achieves 96 ms per one-second spectrogram frame. These results demonstrate a practical path toward privacy-preserving, noise-robust, and communication-efficient infant cry classification suitable for federated deployment.
format Preprint
id arxiv_https___arxiv_org_abs_2512_13880
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Privacy-Enhancing Infant Cry Classification with Federated Transformers and Denoising Regularization
Owino, Geofrey
Shibwabo, Bernard
Machine Learning
Artificial Intelligence
Sound
Infant cry classification can aid early assessment of infant needs. However, deployment of such solutions is limited by privacy concerns around audio data, sensitivity to background noise, and domain shift across recording environments. We present an end-to-end infant cry analysis pipeline that integrates a denoising autoencoder (DAE), a convolutional tokenizer, and a Transformer encoder trained using communication-efficient federated learning (FL). The system performs on-device denoising, adaptive segmentation, post hoc calibration, and energy-based out-of-distribution (OOD) abstention. Federated training employs a regularized control variate update with 8-bit adapter deltas under secure aggregation. Using the Baby Chillanto and Donate-a-Cry datasets with ESC-50 noise overlays, the model achieves a macro F1 score of 0.938, an AUC of 0.962, and an Expected Calibration Error (ECE) of 0.032, while reducing per-round client upload from approximately 36 to 42 MB to 3.3 MB. Real-time edge inference on an NVIDIA Jetson Nano (4 GB, TensorRT FP16) achieves 96 ms per one-second spectrogram frame. These results demonstrate a practical path toward privacy-preserving, noise-robust, and communication-efficient infant cry classification suitable for federated deployment.
title Privacy-Enhancing Infant Cry Classification with Federated Transformers and Denoising Regularization
topic Machine Learning
Artificial Intelligence
Sound
url https://arxiv.org/abs/2512.13880