Deep Recurrent Neural Network for Human Activity Recognition
Masaya Inoue, Sozo Inoue, Takeshi Nishida,
International Symposium on Applied Engineering and Sciences (SAES2016)
In this study, we propose a method of human activity recognition from accelerometer using a deep recurrent neural network (DRNN), which is capable of directly processing time series data, and of estimating activities with high accuracy and high throughput. The “high throughput” refers to short time at a time of recognition. We investigated various parameters and architectures of the DRNN by using the training dataset of 432 trials with 6 activity classes from 7 people. The maximum recognition rate was 95.42% and 83.43% against the test data of 108 segmented trials each of which has single activity class and 18 multiple sequential trials, respectively. Here, the maximum recognition rates by traditional methods were 71.65% and 54.97% for each. In addition, the efficiency of the found parameters was evaluated by using additional dataset. Further, as for throughput of the recognition per unit time, the constructed DRNN was requiring only 1.347 [ms], while the best traditional method required 11.031 [ms] which includes 11.027 [ms] for feature calculation. These advantages have been brought about by the small and simple designed to DRNN. We believe that it can be applied to this network as a real time recognition technology.