ABSTRACT
COVID-19, as a rampant health crisis, lies at the basis of fluctuating perceptions affecting decreased demand among travelers. Recent studies have witnessed a growth of interest in the interactions between tourists' behaviors and other factors with the potential to moderate such behavior during travel. However, it remains to be discussed whether the influence of demographic aspects, especially cultural and gender differences, on tourism behaviors will be more prominent during COVID-19. The current empirical research aims to integrate demographic variables, including gender and culture, with tourists' behavior in terms of their choice of companions, travel destinations, and mode of transportation. According to the research findings, people in other countries have greater desire to travel than Chinese tourists who, in any case, prefer to travel with friends. Tourists from other countries are more willing to travel by plane and by car. Males show a more positive attitude than females to these means of transportation. Moreover, the interactive effect of gender and nationality reveals that female travelers from mainland China put the train or bus top on their agenda. These theoretical findings have the potential to provide actionable insights into how policymakers and service providers can make adjustments to bring back tourism stifled by COVID-19.
ABSTRACT
As the COVID-19 pandemic rampages across the world, the demands of video conferencing surge. To this end, real-time portrait segmentation becomes a popular feature to replace backgrounds of conferencing participants. While feature-rich datasets, models and algorithms have been offered for segmentation that extract body postures from life scenes, portrait segmentation has yet not been well covered in a video conferencing context. To facilitate the progress in this field, we introduce an open-source solution named PP-HumanSeg. This work is the first to construct a large-scale video portrait dataset that contains 291 videos from 23 conference scenes with 14K fine-labeled frames and extensions to multi-camera teleconferencing. Furthermore, we propose a novel Self-supervised Connectivity-aware Learning (SCL) for semantic segmentation, which introduces a self-supervised connectivity-aware loss to improve the quality of segmentation results from the perspective of connectivity. And we propose an ultra-lightweight model with SCL for practical portrait segmentation, which achieves the best trade-off between IoU and the speed of inference. Extensive evaluations on our dataset demonstrate the superiority of SCL and our model.