Saved in:
Bibliographic Details
Main Authors: Deng, Qi, Zhang, Yinghao, Liu, Yalin, Tao, Bishenghui
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.21130
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914111384715264
author Deng, Qi
Zhang, Yinghao
Liu, Yalin
Tao, Bishenghui
author_facet Deng, Qi
Zhang, Yinghao
Liu, Yalin
Tao, Bishenghui
contents Autism Spectrum Disorder (ASD) diagnosis systems in school environments increasingly relies on IoT-enabled cameras, yet pure cloud processing raises privacy and latency concerns while pure edge inference suffers from limited accuracy. We propose Confidence-Constrained Cloud-Edge Knowledge Distillation (C3EKD), a hierarchical framework that performs most inference at the edge and selectively uploads only low-confidence samples to the cloud. The cloud produces temperature-scaled soft labels and distils them back to edge models via a global loss aggregated across participating schools, improving generalization without centralizing raw data. On two public ASD facial-image datasets, the proposed framework achieves a superior accuracy of 87.4\%, demonstrating its potential for scalable deployment in real-world applications.
format Preprint
id arxiv_https___arxiv_org_abs_2510_21130
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Confidence-Constrained Cloud-Edge Collaborative Framework for Autism Spectrum Disorder Diagnosis
Deng, Qi
Zhang, Yinghao
Liu, Yalin
Tao, Bishenghui
Networking and Internet Architecture
Autism Spectrum Disorder (ASD) diagnosis systems in school environments increasingly relies on IoT-enabled cameras, yet pure cloud processing raises privacy and latency concerns while pure edge inference suffers from limited accuracy. We propose Confidence-Constrained Cloud-Edge Knowledge Distillation (C3EKD), a hierarchical framework that performs most inference at the edge and selectively uploads only low-confidence samples to the cloud. The cloud produces temperature-scaled soft labels and distils them back to edge models via a global loss aggregated across participating schools, improving generalization without centralizing raw data. On two public ASD facial-image datasets, the proposed framework achieves a superior accuracy of 87.4\%, demonstrating its potential for scalable deployment in real-world applications.
title A Confidence-Constrained Cloud-Edge Collaborative Framework for Autism Spectrum Disorder Diagnosis
topic Networking and Internet Architecture
url https://arxiv.org/abs/2510.21130