Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Preoperative CT-Based Habitat Radiomics Classifiers Predict Recurrence in Non-Small Cell Lung Cancer.

medRxiv : the preprint server for health sciences·2026

Same author

Large language models robustness against perturbation.

Scientific reports·2025

Same author

Detection of Confounders and Potential Confounders in Computed Tomography Lung Datasets.

Journal of imaging informatics in medicine·2025

Same authorSame journal

Enhancing Neonatal Pain Assessment Transparency via Explanatory Training Examples Identification.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

Same authorSame journal

Enhancing Concept-Based Explanation with Vision-Language Models.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

Same authorSame journal

Automated Deep Learning Approach for Post-Operative Neonatal Pain Detection and Prediction through Physiological Signals.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

Same journal

Prediction of MRI-Induced Power Absorption in Patients with DBS Leads.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

Same journal

The Hidden Threat of Hallucinations in Binary Chest X-ray Pneumonia Classification.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

Same journal

Artificial Intelligence Assurance in Head and Neck Surgery: Now and Next.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems·2025

See all related articles

Search research articles

Related Experiment Video

Updated: Jan 17, 2026

Electrophysiological Measurements and Analysis of Nociception in Human Infants

Electrophysiological Measurements and Analysis of Nociception in Human Infants

Published on: December 20, 2011

Few-Shot Prompting with Vision Language Model for Pain Classification in Infant Cry Sounds.

Anthony McCofie¹, Abhiram Kandiyana¹, Peter R Mouton²

¹Computer Science and Engineering, University of South Florida, Tampa, Florida, USA.

Proceedings. IEEE International Symposium on Computer-Based Medical Systems

|September 22, 2025

Summary

This summary is machine-generated.

Detecting infant pain is challenging. This study uses GPT-4(V) and mel spectrograms for accurate infant pain detection with minimal data, improving interpretability.

Keywords:

few-shot prompting infant pain detection large language model pain classification vision language model

More Related Videos

Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli

Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli

Published on: April 5, 2019

Exploring Infant Sensitivity to Visual Language using Eye Tracking and the Preferential Looking Paradigm

Exploring Infant Sensitivity to Visual Language using Eye Tracking and the Preferential Looking Paradigm

Published on: May 15, 2019

Related Experiment Videos

Last Updated: Jan 17, 2026

Electrophysiological Measurements and Analysis of Nociception in Human Infants

Electrophysiological Measurements and Analysis of Nociception in Human Infants

Published on: December 20, 2011

Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli

Multi-Modal Signals for Analyzing Pain Responses to Thermal and Electrical Stimuli

Published on: April 5, 2019

Exploring Infant Sensitivity to Visual Language using Eye Tracking and the Preferential Looking Paradigm

Exploring Infant Sensitivity to Visual Language using Eye Tracking and the Preferential Looking Paradigm

Published on: May 15, 2019

Area of Science:

Artificial Intelligence
Infant Health
Signal Processing

Background:

Accurate infant pain detection is crucial but challenging.
Conventional deep neural networks require large datasets and computational power, lacking interpretability.
Existing methods struggle with data scarcity and transparency.

Purpose of the Study:

To introduce a novel, interpretable method for infant pain detection using few-shot learning.
To leverage vision-language models (GPT-4(V)) with mel spectrograms for enhanced infant cry analysis.
To reduce the reliance on extensive labeled datasets in infant pain classification.

Main Methods:

Utilized OpenAI's GPT-4(V) vision-language model.
Employed mel spectrogram representations of infant cry sounds.
Implemented a few-shot prompting strategy for classification.
Validated the approach on the USF-MNPAD-II dataset.

Main Results:

Achieved 83.33% accuracy in infant pain detection.
Required only 16 training samples, a significant reduction from the baseline's 4,914 samples.
Demonstrated enhanced transparency and interpretability compared to conventional methods.

Conclusions:

Few-shot prompting with vision-language models offers a promising solution for infant pain detection.
This approach significantly reduces data and computational requirements.
Represents a novel application of GPT-4o for infant pain classification.