Search

Abstract Person image generation aims to generate images that maintain the original human appearance in different target poses. Recent works have revealed that the critical element in achieving this task is the alignment of appearance domain and pose domain. Previous alignment methods, such as appearance flow warping, correspondence learning and cross attention, often encounter challenges when it comes to producing fine texture details. These approaches suffer from limitations in accurately estimating appearance flows due to the lack of global receptive field. Alternatively, they can only perform cross‐domain alignment on high‐level feature maps with small spatial dimensions since the computational complexity increases quadratically with larger feature sizes. In this article, the significance of multi‐scale alignment, in both low‐level and high‐level domains, for ensuring reliable cross‐domain alignment of appearance and pose is demonstrated. To this end, a novel and effective method, named Multi‐scale Cross‐domain Alignment (MCA) is proposed. Firstly, MCA adopts global context aggregation transformer to model multi‐scale interaction between pose and appearance inputs, which employs pair‐wise window‐based cross attention. Furthermore, leveraging the integrated global source information for each target position, MCA applies flexible flow prediction head and point correlation to effectively conduct warping and fusing for final transformed person image generation. Our proposed MCA achieves superior performance on two popular datasets than other methods, which verifies the effectiveness of our approach.

Abstract This article investigates the interplay between knowledge acquisition and the stepwise migration trajectories of low-skilled Vietnamese workers in China. Utilizing a rich array of data garnered from face-to-face interviews, customized surveys, on-site observations, media narratives, and an archival review of government regulations in Guangxi—a province bordering Vietnam—and in Guangdong, a non-adjacent province, this study proposes an analytical framework, skill-driven migration, to highlight the synergy between the pre-migratory socio-demographic attributes of migrants and the circumstances prevailing in both the places of origin and destination which orchestrates distinct job allocations and skill augmentation avenues. Notably, our study underscores that the competencies acquired in manufacturing employments potentially broaden the horizon of destinations for migrants in their subsequent stepwise migration endeavors. This phenomenon, in turn, poses intricate immigration management quandaries for newly emerged destination nations like China, spotlighting a nuanced dimension of labor mobility and policy frameworks in the evolving regional economic integration. Last, our findings challenge the conventional dichotomy of high-skilled and low-skilled categories of labor migration, as well as the dichotomy of explicit-tacit knowledge.

According to the context of “carbon peak and carbon neutralization”, it is highly significant to investigate the melting mechanism and behavior of steel scrap in the converter to reduce the tap-to-tap time and enhance the usage of resource. In this paper, the melting behaviors of steel scrap in a full-scale converter under the conditions of without blowing, only top blowing, only bottom blowing and top-bottom combined blowing were numerically calculated by a three-dimensional turbulent-multiphase-melting-solute transport coupling model. Along with hot-state experiments, the mathematical model was verified. The results showed that the scrap melting in the converter is the result of the combined actions of heat transfer and mass transfer. When applying the top-bottom combined blowing, the melting time required for the scrap is about 485 s. Compared with bottom blowing, top blowing and without blowing, the melting time is shortened by about 45 s, 220 s and 290 s, respectively. The modes of top-bottom combined blowing and only bottom blowing can preferably promote the exchange of temperature and solute element carbon between the scrap and melt, leading to a stable temperature of 1700 K around the scrap, which is 10 K and 20 K higher than that of only top blowing and without blowing, respectively. Meanwhile, the end time of carburizing stage with top-bottom combined blowing is 200 s, which is about 100 s shorter than that of only top blowing and without blowing. The present study can provide theoretical guidance for deep understanding of scrap melting in converter, which is significant for improving the production efficiency.

Introduction: Although the molecular mechanisms of Krüpple-like factor 4 (KLF4) as a tumor suppressor in HCC tumorigenesis have been thoroughly examined, its clinical application in terms of precise prognostication and its influence on tumor immune microenvironment in patients with HCC require further investigation.Methods: Bioinformatics and immunohistochemistry (IHC) were used to validate KLF4 expressions in a tissue microarray (TMA) containing HCC samples. Using Cox regression models, independent prognostic factors were identified and employed in the development of nomograms. Decision curve analysis (DCA) demonstrated the superiority of the nomograms. GO and KEGG pathway analyses were applied to the functional study of KLF4. The GSVA program explored the link between KLF4 expression and tumor-infiltrating immune cells, and CAMOIP was used to construct KLF4 expression immune scores. Changes in immune-related gene markers were also investigated in relation to KLF4 expression. The association between immune cell infiltration and KLF4 expression was validated by IHC in TMA.Results: HCC was reported to have a notable depletion of KLF4. The absence of KLF4 was associated with advanced clinicopathological characteristics of HCC and predicted a bad prognosis for patients. Nomograms constructed using KLF4 expression, tumor differentiation, and TNM stage provided a more accurate prognostic assessment of HCC patients than TNM stage alone. KLF4 expression was associated with immunological-related functions, infiltration of macrophages, CD8+ T cells, and other immune cells, and elevation of immune checkpoints. Higher levels of CD8+ T cells and macrophage infiltration are associated with increased KLF4 expression in HCC TMA.Conclusion: KLF4 loss in HCC is a prognostic biomarker that influences the tumor immune microenvironment (TIME).

Abstract Facial reenactment has developed rapidly in recent years, but few methods have been built upon reenacted face in videos. Facial‐reenacted emotion synthesis can make the process of facial reenactment more practical. A facial‐reenacted emotion synthesis method is proposed that includes a dual‐path generative adversarial network (GAN) for emotion synthesis and a residual‐mask network to impose structural restrictions to preserve the mouth shape of the source person. To train the dual‐path GAN more effectively, a learning strategy based on separated discriminators is proposed. The method is trained and tested on a very challenging imbalanced dataset to evaluate the ability to deal with complex practical scenarios. Compared with general emotion synthesis methods, the proposed method can generate more realistic facial emotion synthesised images or videos with higher quality while retaining the expression contents of the original videos. The DualPathGAN achieves a Fréchet inception distance (FID) score of 9.20, which is lower than the FID score of 11.37 achieved with state‐of‐the‐art methods.

Physical activity is critical for maintaining cognitive and brain health. Previous studies have indicated that the effect of physical activity on cognitive and brain function varies between individuals. The present study aimed to examine whether mind wandering modulated the relations between physical activity and resting-state hippocampal functional connectivity. A total of 99 healthy adults participated in neuroimaging data collection as well as reported their physical activity in the past week and their propensity to mind wandering during typical activities. The results indicated that mind wandering was negatively related to the resting-state functional connectivity between hippocampus and right inferior occipital gyrus. Additionally, for participants with higher level of mind wandering, physical activity was negatively related to hippocampal connectivity at left precuneus and right precentral gyrus. In contrast, such relations were positive at right medial frontal gyrus and bilateral precentral gyrus for participants with lower level of mind wandering. Altogether, these findings indicated that the relations between physical activity and hippocampal functional connectivity vary as a function of mind wandering level, suggesting that individual differences are important to consider when we aim to maintain or improve cognitive and brain health through increasing physical activity.

Abstract Background The stage‐specific roles of radiotherapy (RT) alone, chemotherapy alone, and combined RT and chemotherapy (CRT) for patients with nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) have not been adequately evaluated. Methods We analyzed patients with all stages of NLPHL enrolled in the Surveillance, Epidemiology, and End Results (SEER) registry from January 2000 to December 2015. Propensity score (PS) analysis with 1:1 matching (PSM) was performed to ensure the well‐balanced characteristics of the comparison groups. Kaplan–Meier and Cox proportional‐hazards models were used to evaluate the overall survival (OS), cancer‐specific survival (CSS), hazard ratios (HRs), and corresponding 95% confidence intervals (95% CI). Restricted mean survival times (RMST) were also used for the survival analyses. Results For early‐stage patients, CRT was associated with the best survival, the mean OS was significantly improved by approximately 20 months (20 m), and the risk of death was reduced by more than 80%, both before and after PSM (p