Code-switching is about dealing with alternative languages in speech or text. It is partially speaker-dependent and domainrelated, so completely explaining the phenomenon by linguistic rules is challenging. Compared to most monolingual tasks, insufficient data is an issue for code-switching. To mitigate the issue without expensive human annotation, we proposed an unsupervised method for code-switching data augmentation. By utilizing a generative adversarial network, we can generate intra-sentential code-switching sentences from monolingual sentences. We applied the proposed method on two corpora, and the result shows that the generated code-switching sentences improve the performance of code-switching language models.
The purpose of this study was to compare non-uniform image quality caused by the anode heel effect between two radiographic systems using a circular step-wedge (CSW) phantom and the normalized mutual information (nMI) metric. Ten repeated radiographic images of the CSW and contrast-detail resolution (CDR) phantoms were acquired from two digital radiographic systems with 16- and 12-degree anode angles, respectively, using various kVp and mAs. To compare non-uniform image quality, the CDR phantom was physically rotated at different orientations, and the directional nMI metrics were calculated from the CSW images. The directional visible ratio (VR) metrics were calculated from the CDR images. Analysis of variance (ANOVA) was performed to understand whether the nMI metric significantly changed with kVp, mAs, and orientations with Bonferroni correction. Mann–Whitney’s U test was performed to compare the metrics between the two systems. Contrary to the VR metrics, the nMI metrics significantly changed with orientations in both radiographic systems. In addition, the system with the 12-degree anode angle exhibited less uniform image quality compared to the system with the 16-degree anode angle. A CSW phantom using the directional nMI metric can be significantly helpful to compare non-uniform image quality between two digital radiographic systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.