TDACNN: Target-domain-free Domain Adaptation Convolutional Neural Network for Drift Compensation in Gas Sensors

TL;DR

TDACNN is a novel deep learning model that achieves drift compensation in gas sensors without requiring data from the drifted target domain, using a multi-branch CNN and ensemble methods.

Contribution

It introduces a target-domain-free domain adaptation CNN with a multibranch backbone and ensemble classifier, enabling drift compensation without target domain data during training.

Findings

Outperforms state-of-the-art methods on drift datasets.

Effectively extracts domain-invariant features for drift compensation.

Demonstrates robustness across different drift scenarios.

Abstract

Sensor drift is a long-existing unpredictable problem that deteriorates the performance of gaseous substance recognition, calling for an antidrift domain adaptation algorithm. However, the prerequisite for traditional methods to achieve fine results is to have data from both nondrift distributions (source domain) and drift distributions (target domain) for domain alignment, which is usually unrealistic and unachievable in real-life scenarios. To compensate for this, in this paper, deep learning based on a target-domain-free domain adaptation convolutional neural network (TDACNN) is proposed. The main concept is that CNNs extract not only the domain-specific features of samples but also the domain-invariant features underlying both the source and target domains. Making full use of these various levels of embedding features can lead to comprehensive utilization of different levels of characteristics, thus achieving drift compensation by the extracted intermediate features between two domains. In the TDACNN, a flexible multibranch backbone with a multiclassifier structure is proposed under the guidance of bionics, which utilizes multiple embedding features comprehensively without involving target domain data during training. A classifier ensemble method based on maximum mean discrepancy (MMD) is proposed to evaluate all the classifiers jointly based on the credibility of the pseudolabel. To optimize network training, an additive angular margin softmax loss with parameter dynamic adjustment is utilized. Experiments on two drift datasets under different settings demonstrate the superiority of TDACNN compared with several state-of-the-art methods.