eduzhai > Applied Sciences > Engineering >

ARC-Net Activity Recognition Through Capsules

  • king
  • (0) Download
  • 20210505
  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Human Activity Recognition (HAR) is a challenging problem that needs advancedsolutions than using handcrafted features to achieve a desirable performance.Deep learning has been proposed as a solution to obtain more accurate HARsystems being robust against noise. In this paper, we introduce ARC-Net andpropose the utilization of capsules to fuse the information from multipleinertial measurement units (IMUs) to predict the activity performed by thesubject. We hypothesize that this network will be able to tune out theunnecessary information and will be able to make more accurate decisionsthrough the iterative mechanism embedded in capsule networks. We provideheatmaps of the priors, learned by the network, to visualize the utilization ofeach of the data sources by the trained network. By using the proposed network,we were able to increase the accuracy of the state-of-the-art approaches by 2 .Furthermore, we investigate the directionality of the confusion matrices of ourresults and discuss the specificity of the activities based on the provideddata.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×