PDFed: Privacy-Preserving and Decentralized Asynchronous Federated Learning for Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Balan, Kar, Gilbert, Andrew, Collomosse, John
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916412364161024
author Balan, Kar
Gilbert, Andrew
Collomosse, John
author_facet Balan, Kar
Gilbert, Andrew
Collomosse, John
contents We present PDFed, a decentralized, aggregator-free, and asynchronous federated learning protocol for training image diffusion models using a public blockchain. In general, diffusion models are prone to memorization of training data, raising privacy and ethical concerns (e.g., regurgitation of private training data in generated images). Federated learning (FL) offers a partial solution via collaborative model training across distributed nodes that safeguard local data privacy. PDFed proposes a novel sample-based score that measures the novelty and quality of generated samples, incorporating these into a blockchain-based federated learning protocol that we show reduces private data memorization in the collaboratively trained model. In addition, PDFed enables asynchronous collaboration among participants with varying hardware capabilities, facilitating broader participation. The protocol records the provenance of AI models, improving transparency and auditability, while also considering automated incentive and reward mechanisms for participants. PDFed aims to empower artists and creators by protecting the privacy of creative works and enabling decentralized, peer-to-peer collaboration. The protocol positively impacts the creative economy by opening up novel revenue streams and fostering innovative ways for artists to benefit from their contributions to the AI space.
format Preprint
id arxiv_https___arxiv_org_abs_2409_18245
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PDFed: Privacy-Preserving and Decentralized Asynchronous Federated Learning for Diffusion Models
Balan, Kar
Gilbert, Andrew
Collomosse, John
Image and Video Processing
We present PDFed, a decentralized, aggregator-free, and asynchronous federated learning protocol for training image diffusion models using a public blockchain. In general, diffusion models are prone to memorization of training data, raising privacy and ethical concerns (e.g., regurgitation of private training data in generated images). Federated learning (FL) offers a partial solution via collaborative model training across distributed nodes that safeguard local data privacy. PDFed proposes a novel sample-based score that measures the novelty and quality of generated samples, incorporating these into a blockchain-based federated learning protocol that we show reduces private data memorization in the collaboratively trained model. In addition, PDFed enables asynchronous collaboration among participants with varying hardware capabilities, facilitating broader participation. The protocol records the provenance of AI models, improving transparency and auditability, while also considering automated incentive and reward mechanisms for participants. PDFed aims to empower artists and creators by protecting the privacy of creative works and enabling decentralized, peer-to-peer collaboration. The protocol positively impacts the creative economy by opening up novel revenue streams and fostering innovative ways for artists to benefit from their contributions to the AI space.
title PDFed: Privacy-Preserving and Decentralized Asynchronous Federated Learning for Diffusion Models
topic Image and Video Processing
url https://arxiv.org/abs/2409.18245