Partial Domain Adaptation Using Selective Representation Learning For Class-Weight Computation

TL;DR

This paper proposes a novel partial domain adaptation method that identifies outlier classes using image content and clusters class weights to improve transfer learning when source and target label sets differ.

Contribution

It introduces a selective representation learning approach that detects outlier classes and clusters class weights to reduce negative transfer in partial domain adaptation.

Findings

Effective identification of outlier classes from image content.

Improved adaptation performance by clustering class weights.

Reduction of negative transfer from source-private classes.

Abstract

The generalization power of deep-learning models is dependent on rich-labelled data. This supervision using large-scaled annotated information is restrictive in most real-world scenarios where data collection and their annotation involve huge cost. Various domain adaptation techniques exist in literature that bridge this distribution discrepancy. However, a majority of these models require the label sets of both the domains to be identical. To tackle a more practical and challenging scenario, we formulate the problem statement from a partial domain adaptation perspective, where the source label set is a super set of the target label set. Driven by the motivation that image styles are private to each domain, in this work, we develop a method that identifies outlier classes exclusively from image content information and train a label classifier exclusively on class-content from source images. Additionally, elimination of negative transfer of samples from classes private to the source domain is achieved by transforming the soft class-level weights into two clusters, 0 (outlier source classes) and 1 (shared classes) by maximizing the between-cluster variance between them.