Source Themes

SleepPoseNet: multi-view learning for sleep postural transition recognition using UWB
We proposed a non-contact sleep posture monitoring system using UWB radar. However, the signals corresponding to sleep postures received by a single antenna are similar among classes, resulting in a low classification accuracy. We created a deep learning model that combine time and time-frequency representations to tackle this problem. Our algorithm improves performance of classification accuracy when the signals are similar among classes
Maytus Piriyajitakonkij*, Patchanon Warin, Payongkit Lakhan, Pitshaporn Leelaarporn, Nakorn Kumchaiseemak, Supasorn Suwajanakorn, Theerasarn Pianpanit, Nattee Niparnan, Subhas Chandra Mukhopadhyay, Theerawit Wilaiprasitporn
PDF Cite Code Slides
SleepPoseNet: multi-view learning for sleep postural transition recognition using UWB
EEGANet: removal of ocular artifact from the EEG signal using generative adversarial networks
The elimination of ocular artifacts is critical in analyzing electroencephalography (EEG) data for various brain-computer interface …
Phattarapong Sawangjai, Manatsanan Trakulruangroj, Chiraphat Boonnag, Maytus Piriyajitakonkij*, Rajesh Kumar Tripathy, Thapanun Sudhawiyangkul, Theerawit Wilaiprasitporn
PDF Code Dataset Poster Video Source Document Custom Link
EEGANet: removal of ocular artifact from the EEG signal using generative adversarial networks