Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

kennygeemartin/Privacy-Preserving-FL-MobileViT

Domain:

healthcare

Record type:

datasetmodel
Creator:
ken
Host:
This study presents an application-oriented privacy-preserving federated learning framework integrating Mobile Vision Transformer (MobileViT) for chest X-ray classification across simulated Nigerian healthcare environments. Using the Nigeria Chest X-ray Dataset (2,600 radiologist-labelled images across four classes) # Privacy-preserving federated MobileViT for Nigerian chest X-rays This repository downloads the public Kaggle dataset, creates reproducible 70/15/15 stratified splits, partitions the training set across five simulated hospitals with a Dirichlet distribution, trains MobileViT-XS using FedAvg, and exports an inference-ready model plus evaluation results. ## Quick start ```powershell python -m pip install -r requirements.txt python train.py download python train.py inspect python train.py train --mode federated ``` The full paper configuration is in `config.yaml`. For a quick end-to-end check: ```powershell python train.py train --mode federated --rounds 1 --local-epochs 1 --max-images 80 --no-pretrained ``` Other experiment modes are `centralized` and `local`. Results are written under `artifacts/ /`, including `best_model.pt`, `model_scripted.pt`, `metrics.json`, `classification_report.csv`, `confusion_matrix.csv`, split manifests, client partitions, and training history. > This is research software, not a medical device. Predictions require clinical > validation and must not be used as a substitute for qualified diagnosis. ## System utility interface The dashboard and model-serving application are in `system-utility-interface`. It reproduces the five-hospital FL workflow and connects uploaded X-rays to the latest exported MobileViT model. See `system-utility-interface/README.md` for launch instructions.

Visit

github.com

Tasks

image classificationcomputer vision

Similar

Federated Learning and Differential Privacy: Software tools analysis, the Sherpa.ai FL framework and methodological guidelines for preserving data privacyPrivacy-Preserving Federated Learning for Hate Speech DetectionVectorgram: An Offline, Privacy-Preserving, Breast Cancer Diagnosis PlatformEvaluation of Privacy Preserving Protocols for Mobile Money - DatasetFederated Learning Enables Privacy-Preserving Cross-Border Malaria SurveillanceDesigning a Privacy-Preserving Behavioral Phishing Detection Artifact: A Design Science Framework Using Federated Differential Privacy

Federated Learning and Differential Privacy: Software tools analysis, the Sherpa.ai FL framework and methodological guidelines for preserving data privacy

The high demand of artificial intelligence services at the edges that also preserve data privacy has

Privacy-Preserving Federated Learning for Hate Speech Detection

This paper presents a federated learning system with differential privacy for hate speech detection,

Vectorgram: An Offline, Privacy-Preserving, Breast Cancer Diagnosis Platform

Vectorgram: An Offline, Privacy-Preserving, Breast Cancer Diagnosis Platform

Poster presented at the Deep Learning Indaba 2023 by Ian Omung’a

Evaluation of Privacy Preserving Protocols for Mobile Money - Dataset

The files herein represent qualitative interviews conducted with users (both customers

Federated Learning Enables Privacy-Preserving Cross-Border Malaria Surveillance

Malaria surveillance in West Africa faces two conflicting imperatives:

    Designing a Privacy-Preserving Behavioral Phishing Detection Artifact: A Design Science Framework Using Federated Differential Privacy

    Current detection frameworks largely analyze emails as technical artifacts while overlooking the beh