Geodesic Kernel embedding Distribution Alignment for domain adaptation

Abstract

Domain adaptation is a method to classify the new domain accurately by using the marked image of the old domain. It shows a good but a challenging application prospect in computer vision. In this article, we propose a unified and optimized problem modeling method, which is called as Geodesic Kernel embedding Distribution Alignment (GKDA). Specifically, GKDA aims to reduce the domain differences. GKDA avoids degenerated feature transformation by using geodesic kernel mapping feature, and then adjusts the weight of cross-domain instances in the process of dimensionality reduction in principle, finally, constructs a new feature to represent the difference of distribution and unrelated instances. The experiment result shows that GKDA has obvious superiority in cross-domain image recognition.

Improved Semantic Representation Learning by Multiple Clustering for Image-Based 3D Model Retrieval

“…Shen et al (2018) computed and minimized the Wasserstein distance between source and target domains. Moreover, a series of new discrepancy measurement criteria are emerged in (Long et al, 2017;Pan & Yang, 2020;Si et al, 2021;Yan et al, 2017). The latter domain adversarial confusion-based methods additionally add a domain discriminator to distinguish the source samples from target samples while the feature extractor is fooled to generate a domain-invariant feature.…”