Exploiting Convolution Filter Patterns for Transfer Learning

TL;DR

This paper presents a novel regularization method for transfer learning that captures statistical relationships among convolution filters using Gaussian Mixture Models, improving transferability across different CNN architectures.

Contribution

The paper introduces a GMM-based regularization technique that leverages filter pattern correlations for enhanced transfer learning in CNNs, independent of architecture.

Findings

Improves transfer learning efficiency and effectiveness

Demonstrates generalizability across datasets and models

Applicable to various CNN architectures

Abstract

In this paper, we introduce a new regularization technique for transfer learning. The aim of the proposed approach is to capture statistical relationships among convolution filters learned from a well-trained network and transfer this knowledge to another network. Since convolution filters of the prevalent deep Convolutional Neural Network (CNN) models share a number of similar patterns, in order to speed up the learning procedure, we capture such correlations by Gaussian Mixture Models (GMMs) and transfer them using a regularization term. We have conducted extensive experiments on the CIFAR10, Places2, and CMPlaces datasets to assess generalizability, task transferability, and cross-model transferability of the proposed approach, respectively. The experimental results show that the feature representations have efficiently been learned and transferred through the proposed statistical regularization scheme. Moreover, our method is an architecture independent approach, which is applicable for a variety of CNN architectures.