Recently, many attention-based deep neural networks have emerged and achieved state-of-the-art performance in environmental sound classification. The essence of attention mechanism is assigning contribution weights on different parts of features, namely channels, spectral or spatial contents, and temporal frames. In this paper, we propose an effective convolutional neural network structure with a multichannel temporal attention (MCTA) block, which applies a temporal attention mechanism within each channel of the embedded features to extract channel-wise relevant temporal information. This multi-channel temporal attention structure will result in a distinct attention vector for each channel, which enables the network to fully exploit the relevant temporal information in different channels. The datasets used to test our model include ESC-50 and its subset ESC-10, along with development sets of DCASE 2018 and 2019. In our experiments, MCTA performed better than the single-channel temporal attention model and the non-attention model with the same number of parameters. Furthermore, we compared our model with some successful attention-based models and obtained competitive results with a relatively lighter network.

The long-term mortality risk of natural disasters is a key threat to disaster resilience improvement, yet an authoritative certification and a reliable surveillance system are, unfortunately, yet to be established in many countries. This study aimed to clarify the mechanism of post-disaster indirect deaths in Japan, to improve the existing disaster recovery evaluation system and support decision making in public policy. This study first investigated the definition of indirect deaths via a literature review before examining the observed number of indirect deaths via case study, census data from the Population Demographic and Household Surveys, other social surveys, and reports in the case of the Great East Japan Earthquake and Tsunami, which severely damaged northeastern Japan, especially the three prefectures, which are the target areas in this context (i.e., Fukushima, Iwate, and Miyagi). It was found that the reported number of indirect deaths was significantly underestimated. In total, 4657 indirect deaths were estimated to have occurred in the target prefectures. This was higher than the reported number, which was 3784. The overall statistics established via collaboration between local administrations and governments can be improved to provide better reference for researchers and policymakers to investigate the long-term effects of natural disaster.

No abstract

No abstract

No abstract