Enregistré dans:
Détails bibliographiques
Auteurs principaux: Tripurwar, Chinmay, Maurya, Utkarsh, Dishant
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2511.20702
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866914171216461824
author Tripurwar, Chinmay
Maurya, Utkarsh
Dishant
author_facet Tripurwar, Chinmay
Maurya, Utkarsh
Dishant
contents Model pruning is a widely adopted technique to reduce the computational complexity and memory footprint of Deep Neural Networks (DNNs). However, global unstructured pruning often leads to significant degradation in accuracy, typically necessitating fine-tuning on the original training dataset to recover performance. In privacy-sensitive domains such as healthcare or finance, access to the original training data is often restricted post-deployment due to regulations (e.g., GDPR, HIPAA). This paper proposes a Data-Free Knowledge Distillation framework to bridge the gap between model compression and data privacy. We utilize DeepInversion to synthesize privacy-preserving ``dream'' images from the pre-trained teacher model by inverting Batch Normalization (BN) statistics. These synthetic images serve as a transfer set to distill knowledge from the original teacher to the pruned student network. Experimental results on CIFAR-10 across various architectures (ResNet, MobileNet, VGG) demonstrate that our method significantly recovers accuracy lost during pruning without accessing a single real data point.
format Preprint
id arxiv_https___arxiv_org_abs_2511_20702
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Post-Pruning Accuracy Recovery via Data-Free Knowledge Distillation
Tripurwar, Chinmay
Maurya, Utkarsh
Dishant
Machine Learning
Artificial Intelligence
Model pruning is a widely adopted technique to reduce the computational complexity and memory footprint of Deep Neural Networks (DNNs). However, global unstructured pruning often leads to significant degradation in accuracy, typically necessitating fine-tuning on the original training dataset to recover performance. In privacy-sensitive domains such as healthcare or finance, access to the original training data is often restricted post-deployment due to regulations (e.g., GDPR, HIPAA). This paper proposes a Data-Free Knowledge Distillation framework to bridge the gap between model compression and data privacy. We utilize DeepInversion to synthesize privacy-preserving ``dream'' images from the pre-trained teacher model by inverting Batch Normalization (BN) statistics. These synthetic images serve as a transfer set to distill knowledge from the original teacher to the pruned student network. Experimental results on CIFAR-10 across various architectures (ResNet, MobileNet, VGG) demonstrate that our method significantly recovers accuracy lost during pruning without accessing a single real data point.
title Post-Pruning Accuracy Recovery via Data-Free Knowledge Distillation
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2511.20702