Disentangled Adversarial Autoencoder for Subject-Invariant Physiological Feature Extraction
This work addresses the problem of user and task variability in physiological sensing for applications like device manipulation, representing an incremental advancement in domain-specific feature extraction.
The paper tackled the challenge of biosignal variability across users and tasks by proposing an adversarial feature extractor for transfer learning, resulting in up to 8.8% improvement in average classification accuracy in cross-subject evaluations.
Recent developments in biosignal processing have enabled users to exploit their physiological status for manipulating devices in a reliable and safe manner. One major challenge of physiological sensing lies in the variability of biosignals across different users and tasks. To address this issue, we propose an adversarial feature extractor for transfer learning to exploit disentangled universal representations. We consider the trade-off between task-relevant features and user-discriminative information by introducing additional adversary and nuisance networks in order to manipulate the latent representations such that the learned feature extractor is applicable to unknown users and various tasks. Results on cross-subject transfer evaluations exhibit the benefits of the proposed framework, with up to 8.8% improvement in average accuracy of classification, and demonstrate adaptability to a broader range of subjects.