Sequential Deep Learning for Action Recognition with Synthetic Multi-view Data from Depth Maps.

Bin Liang Lihong Zheng Xinying Li

Published in: AusDM (2018)

Keyphrases

multi view
depth map
action recognition
data sets
three dimensional
d scene
d objects
multi view stereo
depth data
view synthesis
single view
semi supervised
multiple views
stereo matching
depth information
high quality
multiple cameras
data points
input data
depth images
image processing
machine learning