International Journal of Disaster Risk Reduction

Researchers at The University of Osaka developed a deep learning model for rapid building damage assessment after floods using satellite imagery. This research establishes the first systematic benchmark for this task and introduces a novel semi-supervised learning method achieving 74% of fully supervised performance with just 10% of the labeled data. A new, lightweight deep learning model named Simple Prior Attention Disaster Assessment Net or SPADANet significantly reduces missed damaged buildings, improving recall by over 9% compared to existing models. This work provides crucial design principles for future AI disaster response, enabling faster and more efficient life-saving operations.

In a study on public psychology with regard to the COVID-19 pandemic, researchers from Osaka University found no significant differences among regions of Japan regarding either risk perception or infection-prevention behaviors during the “state of emergency,” suggesting spillover effects between targeted and non-targeted regions. However, risk perception diminished after restrictions were lifted, and both risk perception and hygienic behaviors saw further reductions after the downgrading of the legal status of COVID-19 to that of a common seasonal flu.

Researchers from Osaka University found that Japanese people who used medical information were more likely to wear masks and disinfect their hands even before the COVID-19 outbreak, while those who exchanged information with acquaintances were more likely to newly start mask-wearing and hand disinfection post outbreak. This will assist in the development of risk communication strategies to promote infection-preventive behavior in a timely manner.

A researcher from Osaka University has investigated the associations between mask usage and the reasons people give for wearing a mask. Online surveys were conducted before and after policies that downgraded the status of COVID-19. The study found that 59% of respondents continued wearing masks even after policy changes. Several sociological and psychological reasons other than disease prevention for mask usage were identified.

A study by Osaka University has shown that the prevalence of non-communicable diseases, which included hypertension, hyperlipidemia, diabetes, and mental disorders, increased after the Fukushima disaster and the COVID-19 outbreak. These findings emphasize the importance of improving post-disaster health promotion strategies and recommendations.