Targeted adversarial discriminative domain adaptation

Abstract

Domain adaptation is a technology enabling aided target recognition and other algorithms for environments and targets with data or labeled data that is scarce. Recent advances in unsupervised domain adaptation have demonstrated excellent performance but only when the domain shift is relatively small. We proposed targeted adversarial discriminative domain adaptation (T-ADDA), a semi-supervised domain adaptation method that extends the ADDA framework. By providing at least one labeled target image per class, used as a cue to guide the adaption, T-ADDA significantly boosts the performance of ADDA and is applicable to the challenging scenario in which the sets of targets in the source and target domains are not the same. The efficacy of T-ADDA is demonstrated by cross-domain, cross-sensor, and cross-target experiments using the common digits datasets and several aerial image datasets. Results demonstrate an average increase of 15% improvement with T-ADDA over ADDA using just a few labeled images when adapting to a small domain shift and afforded a 60% improvement when adapting to large domain shifts. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

Value-based sensor and information fusion

“…Examples exist among domains such as ground-based robotics [63] and air-based systems [64]. The collaborative networks should support all-domain mediums of operations such as space [65,66], air, ground, sea, and cyber [67], and support domain adaptation [68,69,70] of DL methods.…”

Digital twin meets information fusion: panel summary

“…One example from multi-modal image fusion [79,80] includes coordinating synthetic aperture radar (SAR) [81,82] and electro-optical (EO) imagery [83]. Using the analysis from one domain (e.g., EO) can be thought of as an automatic target recognition (ATR) twin to support domain adaptation to another domain (e.g., SAR) [84,85,86,87]. The type of ATR-twin to support image translation [88] or domain adaptation supports adaptive learning [89] and responses operating conditions such as a change in the weather [90].…”