Diversify and Match: A Domain Adaptive Representation Learning Paradigm for Object Detection
Authors:
- 1 June 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 12448-12457
Abstract
We introduce a novel unsupervised domain adaptation approach for object detection. We aim to alleviate the imperfect translation problem of pixel-level adaptations, and the source-biased discriminativity problem of feature-level adaptations simultaneously. Our approach is composed of two stages, i.e., Domain Diversification (DD) and Multi-domain-invariant Representation Learning (MRL). At the DD stage, we diversify the distribution of the labeled data by generating various distinctive shifted domains from the source domain. At the MRL stage, we apply adversarial learning with a multi-domain discriminator to encourage feature to be indistinguishable among the domains. DD addresses the source-biased discriminativity, while MRL mitigates the imperfect image translation. We construct a structured domain adaptation framework for our learning paradigm and introduce a practical way of DD for implementation. Our method outperforms the state-of-the-art methods by a large margin of 3%~11% in terms of mean average precision (mAP) on various datasets.Keywords
All Related Versions
This publication has 22 references indexed in Scilit:
- Extreme Clicking for Efficient Object Annotation2017
- Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks2017
- Iterative object and part transfer for fine-grained recognition2017
- Image-to-Image Translation with Conditional Adversarial Networks2017
- The Cityscapes Dataset for Semantic Urban Scene Understanding2016
- Fast R-CNN2015
- Transfer Sparse Coding for Robust Image Representation2013
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009
- Rapid object detection using a boosted cascade of simple features2005
- Histograms of Oriented Gradients for Human Detection2005