M. Ragab et al., "Adversarial Multiple-Target Domain Adaptation for Fault Classification," in IEEE Transactions on Instrumentation and Measurement, doi: 10.1109/TIM.2020.3009341.
Abstract:
Data-driven fault classification methods are receiving great attention as they can be applied to many real-world applications. However, they work under the assumption that training data and testing data are drawn from the same distribution. Practical scenarios have varying operating conditions, which results in a domain shift problem that significantly deteriorate the diagnosis performance. Recently domain adaptation has been explored to address the domain shift problem by transferring the knowledge from labeled source domain (e.g., source working condition) to unlabeled target domain (e.g., target working condition). Yet all the existing methods are working under single source single target (1S1T) settings. Hence, a new model need to be trained for each new target domain. This shows limited scalability in handling multiple working conditions since different models should be trained for different target working conditions, which is clearly not a viable solution in practice. To address this problem, we propose a novel \textit{adversarial multiple domain adaptation} (AMDA) method for single source multiple target (1SmT) scenario, where the model can generalize to multiple target domains concurrently. Adversarial adaptation is applied to transform the multiple target domains features to be invariant from the single source domain features. This leads to a scalable model with a novel capability of generalizing to multiple target domains. Extensive experiments on two public datasets and one self-collected dataset have demonstrated that the proposed method outperforms state-of-the-art methods consistently. Our source codes and data are available via https://github.com/mohamedr002/AMDA.
License type:
Funding Info:
This work is supported by the A*STAR Industrial Internet of Things Research Program under the RIE2020 IAF-PP Grant (grant no: A1788a0023), and partially supported by National Natural Science Foundation of China (No. 51835009). In addition, the first author of this work is supported by A*STAR SINGA Scholarship.