Search

We analyze the GeV $\gamma$-ray emission from the W3 complex using about 14 years of Pass 8 data recorded by the $\it Fermi$ Large Area Telescope (\textit{Fermi}-LAT). We resolve the $\gamma$-ray emissions around W3 into two components: an elliptical Gaussian overlapping with the molecular gas and a point-like source near the cluster W3 Main. The pion-bump feature of SED for the elliptical Gaussian together with the better fitting result of pion decay model favor the hadronic origin. We further argue that the cosmic rays (CRs) could originate from the interactions between cluster winds and the shock produced by the SNR HB3. The point-like source positionally coincident with the star cluster W3 Main indicates it may be directly powered by near clusters, while its fainter $\gamma$-ray emissions below 10 GeV is possibly due to the shelter from dense gas making the low-energy CRs incapable of penetrating the dense materials. Meanwhile, we cannot rule out that the $\gamma$-ray emissions originate from the interaction of accelerated protons in SNR with the ambient gas.

Unsupervised Domain Adaptation (UDA) approaches address the covariate shift problem by minimizing the distribution discrepancy between the source and target domains, assuming that the label distribution is invariant across domains. However, in the imbalanced domain adaptation (IDA) scenario, covariate and long-tailed label shifts both exist across domains. To tackle the IDA problem, some current research focus on minimizing the distribution discrepancies of each corresponding class between source and target domains. Such methods rely much on the reliable pseudo labels' selection and the feature distributions estimation for target domain, and the minority classes with limited numbers makes the estimations more uncertainty, which influences the model's performance. In this paper, we propose a cross-domain class discrepancy minimization method based on accumulative class-centroids for IDA (centroIDA). Firstly, class-based re-sampling strategy is used to obtain an unbiased classifier on source domain. Secondly, the accumulative class-centroids alignment loss is proposed for iterative class-centroids alignment across domains. Finally, class-wise feature alignment loss is used to optimize the feature representation for a robust classification boundary. A series of experiments have proved that our method outperforms other SOTA methods on IDA problem, especially with the increasing degree of label shift.