主催: 電気・情報関係学会九州支部連合大会委員会
会議名: 2019年度電気・情報関係学会九州支部連合大会
回次: 72
開催地: 九州工業大学
開催日: 2019/09/27 - 2019/09/28
This paper presents the detection of basic human movements in interaction with different objects using deep learning. The goal is to retrieve the coordinates of the hand and the interacting object, classify these movements and train them to obtain a model capable of correct prediction. The test program uses a video input and by detecting the coordinates of the hand and the object on a new video, the model is predicting which movement is done on the video and build sentences about all the movement present in it.