Abstract
Domain adaptation (DA), a particular case of transfer learning, is an effective technology for learning a discriminative model in scenarios where the data from the training (source) and the testing (target) domains share common class labels but follow different distributions. The differences between domains, called domain shifts, are caused by variations in the acquisition devices and environmental conditions, such as changing illuminations, pose, and collecting-device noises, that are related to a specific domain, denoted as domain-specific noises in this paper. The research on stacked denoising autoencoder (SDA) has demonstrated that noise-robust features could be learned through training a model to reduce the man-made (simulated) noises. However, little research has been conducted to learn the domain-invariant features through training SDA to reduce the domain-specific noises from the real word. In this paper, we propose a novel variant of SDA for DA, called the stacked local constraint auto-encoder (SLC-AE), which aims to learn domain-invariant features through iteratively optimizing the SDA and the low-dimensional manifold. The core idea behind the SLC-AE is that both the source and target samples are corrupted due to the domain-specific noises, and each corrupted sample could be de-noised by calculating the weighted sum of its neighbor samples defined on the intrinsic manifold. Because the neighbor samples on the intrinsic manifold are semantically similar, their weighted sum preserves the generic information and reduces the domain-specific noises. To properly evaluate the performance of the SLC-AE, we conducted extensive experiments using seven benchmark data sets, i.e., MNIST, USPS, COIL20, SYN SIGNS, GTSRB, MSRC and VOC 2007. Compared to twelve different state-of-the-art methods, the experimental results demonstrated that the proposed SLC-AE model made significant improvement over the performance of SDA and achieved the best average performance on the seven data sets.
Highlights
Supervised learning with deep architectures has made remarkable contributions to machine learning and computer vision, leading to the development of robust algorithms that are applicable to a broad range of application problems
EXPERIMENTS we evaluate the performance of the proposed stacked local constraint auto-encoder (SLC–AE) on seven benchmark data sets, i.e., COIL20 [47], MNIST, USPS, SYN Signs [48], GTSRB [49], VOC 2007 [50] and MSRC [51]
In this paper, we present our research in Domain adaptation (DA) with a focus in an area that assumes that the originally collected data in different domains are ‘corrupted’ by the domain-specific noises from the real world
Summary
Supervised learning with deep architectures has made remarkable contributions to machine learning and computer vision, leading to the development of robust algorithms that are applicable to a broad range of application problems. Usually follow different distributions or underlying structures In such scenarios, the performance of the conventional machine learning models is significantly decreased on the testing data, even though the labeled training data are large. In the case of handwritten digit recognition, the support vector machine (SVM) model, a conventional machine learning model, trained using the training data from the USPS data set, can achieve about 88.7% accuracy on the testing data from the same data set [1] It can achieve only about 33.2% accuracy on the testing data from the MNIST data set [2], which shares the 10 same classes of digits but follows different feature distributions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
