BP-Triplet Net for Unsupervised Domain Adaptation: A Bayesian Perspective

TL;DR

This paper introduces BP-Triplet Net, a Bayesian-based triplet loss method for unsupervised domain adaptation that emphasizes hard pairs and improves domain alignment, leading to better target error bounds.

Contribution

It proposes a novel BP-Triplet loss that weights pair-wise samples based on Bayesian principles, enhancing feature learning and domain alignment in UDA.

Findings

Achieved low joint error on multiple benchmarks.

Effectively self-attends to hard positive and negative pairs.

Improves target pseudo label quality progressively.

Abstract

Triplet loss, one of the deep metric learning (DML) methods, is to learn the embeddings where examples from the same class are closer than examples from different classes. Motivated by DML, we propose an effective BP-Triplet Loss for unsupervised domain adaption (UDA) from the perspective of Bayesian learning and we name the model as BP-Triplet Net. In previous metric learning based methods for UDA, sample pairs across domains are treated equally, which is not appropriate due to the domain bias. In our work, considering the different importance of pair-wise samples for both feature learning and domain alignment, we deduce our BP-Triplet loss for effective UDA from the perspective of Bayesian learning. Our BP-Triplet loss adjusts the weights of pair-wise samples in intra domain and inter domain. Especially, it can self attend to the hard pairs (including hard positive pair and hard negative pair). Together with the commonly used adversarial loss for domain alignment, the quality of target pseudo labels is progressively improved. Our method achieved low joint error of the ideal source and target hypothesis. The expected target error can then be upper bounded following Ben-David s theorem. Comprehensive evaluations on five benchmark datasets, handwritten digits, Office31, ImageCLEF-DA, Office-Home and VisDA-2017 demonstrate the effectiveness of the proposed approach for UDA.