Evaluation of Deep Neural Network Domain Adaptation Techniques for Image Recognition

TL;DR

This paper evaluates four unsupervised deep domain adaptation techniques for image recognition, analyzing their effectiveness in improving model generalization across different datasets.

Contribution

It provides a comparative analysis of DeepCORAL, DeepDomainConfusion, CDAN, and CDAN+E on the office-31 dataset, highlighting their relative performance.

Findings

DeepCORAL and CDAN+E outperform other methods in accuracy.

Unsupervised domain adaptation improves generalization to target datasets.

The report offers a practical implementation on GitHub.

Abstract

It has been well proved that deep networks are efficient at extracting features from a given (source) labeled dataset. However, it is not always the case that they can generalize well to other (target) datasets which very often have a different underlying distribution. In this report, we evaluate four different domain adaptation techniques for image classification tasks: DeepCORAL, DeepDomainConfusion, CDAN and CDAN+E. These techniques are unsupervised given that the target dataset dopes not carry any labels during training phase. We evaluate model performance on the office-31 dataset. A link to the github repository of this report can be found here: https://github.com/agrija9/Deep-Unsupervised-Domain-Adaptation.