Unveiling Hidden Patterns in Infant Cry Audio: A Multi-Feature Vision Transformer Approach With Explainable AI
Bibliographic record
Abstract
The early detection and diagnosis of neonatal problems are critical to ensuring that an infant receives timely medical attention, which greatly enhances health outcomes. In this study, we propose a novel deep learning framework that listens to an infant’s cry to identify and diagnose six separate conditions: one being healthy and the other five comprising sepsis, respiratory distress syndrome, jaundice, hyperbilirubinemia, and vomiting. The study utilizes a rich dataset of infant cry recordings from which key acoustic features such as spectrograms, Mel-spectrograms, and Gammatone Frequency Cepstral Coefficients (GFCCs) are extracted. A sophisticated Vision Transformer (ViT) model was developed and meticulously fine-tuned to achieve an impressive 99% classification accuracy through cross-validation. To enhance the model’s interpretability, powerful explainable artificial intelligence (XAI) methods such as LRP, LIME, and attention imaging were implemented to clarify the reasoning behind the model’s outputs. Through cross-validation tests, the model’s trustworthiness and extensive generalizability were assessed. The findings underscore the promising capabilities of employing transformer-based deep learning frameworks along with multimodal acoustic features and explanatory methods to improve cry analysis in infants and their usable scopes in pediatric medicine.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".