Improving Transferability of Deep Neural Networks

TL;DR

This paper investigates how selecting optimal learning rates for neural network layers in transfer learning significantly improves accuracy, demonstrated through experiments on image datasets and real-world tasks.

Contribution

It introduces a method to determine optimal layer-specific learning rates based on dataset parameters, enhancing transfer learning performance.

Findings

Accuracy improvements of up to 127% with proper learning rate selection

Dataset image/label ratio can guide optimal learning rate choice

Validated method on real-world image classification tasks

Abstract

Learning from small amounts of labeled data is a challenge in the area of deep learning. This is currently addressed by Transfer Learning where one learns the small data set as a transfer task from a larger source dataset. Transfer Learning can deliver higher accuracy if the hyperparameters and source dataset are chosen well. One of the important parameters is the learning rate for the layers of the neural network. We show through experiments on the ImageNet22k and Oxford Flowers datasets that improvements in accuracy in range of 127% can be obtained by proper choice of learning rates. We also show that the images/label parameter for a dataset can potentially be used to determine optimal learning rates for the layers to get the best overall accuracy. We additionally validate this method on a sample of real-world image classification tasks from a public visual recognition API.