Multi-Modality Tensor Fusion Based Human Fatigue Detection

Multimodal learning is an expanding research area and aims to pursue a better understanding of given data by regarding different modals. approaches for qualitative are used the quantitative proofing ground-truth datasets discovering unexpected phenomena. In this paper, we investigate effect multimodal schemes assess its state. We try interpret human fatigue levels through analyzing video, thermal image voice together. The experiment showed that approach using three types was more effective than method each dataset individually. As result, identified possibility predicting states.