The recognition of human activity is a task that is applicable to various domains, such as health care, preventive medicine, and elderly care. Furthermore, with the rapid popularization of devices equipped with built-in sensors such as smartphones in recent years, the detection cost of devices has drastically decreased. As a result, researches on mobile activity recognition have been actively conducted.
In our laboratory, we have realized high-performance and high-throughput recognition by using Deep Recurrent Neural Network (DRNN) which is a deep learning method which is good at processing time-series signals.

★★Preprint★★
(●ﾟ◇ﾟ●)Published source code(●ﾟ◇ﾟ●)