Maximum Classifier Discrepancy for Unsupervised Domain Adaptation

Code

• github.com/mil-tokyo/MCD_DA
  OfficialIn paperpytorch★ 0
• github.com/thuml/Transfer-Learning-Library
  pytorch★ 3,889
• github.com/kevinmusgrave/pytorch-adapt
  pytorch★ 394
• github.com/adapt-python/adapt
  tf★ 366
• github.com/mcd4874/neurips_competition
  pytorch★ 28
• github.com/Nyn-ynu/MCD
  pytorch★ 0
• github.com/onedayatatime0923/Cycle_Mcd_Gan
  pytorch★ 0
• github.com/HiGal/Domain-Adaptation
  pytorch★ 0
• github.com/Elman295/Maximum-Classifier-Discrepancy-for-Unsupervised-Domain-Adaptation
  none★ 0

Abstract

In this work, we present a method for unsupervised domain adaptation. Many adversarial learning methods train domain classifier networks to distinguish the features as either a source or target and train a feature generator network to mimic the discriminator. Two problems exist with these methods. First, the domain classifier only tries to distinguish the features as a source or target and thus does not consider task-specific decision boundaries between classes. Therefore, a trained generator can generate ambiguous features near class boundaries. Second, these methods aim to completely match the feature distributions between different domains, which is difficult because of each domain's characteristics. To solve these problems, we introduce a new approach that attempts to align distributions of source and target by utilizing the task-specific decision boundaries. We propose to maximize the discrepancy between two classifiers' outputs to detect target samples that are far from the support of the source. A feature generator learns to generate target features near the support to minimize the discrepancy. Our method outperforms other methods on several datasets of image classification and semantic segmentation. The codes are available at https://github.com/mil-tokyo/MCD_DA

Tasks

Benchmark Results

Reproductions