View-invariant 3D Skeleton-based Human Activity Recognition based on Transformer and Spatio-temporal Features.

Ahmed Snoun Tahani Bouchrika Olfa Jemai

Published in: ICPRAM (2022)

Keyphrases

human activity recognition
view invariant
activity recognition
action recognition
human activities
human actions

multi view
single view
viewpoint
human motion
visual surveillance
spatio temporal

human detection
human computer interaction
sensor data
computer vision
semi supervised
human pose

multiple views
body parts
space time
labeled data
three dimensional
feature selection