PMID- 36090467
OWN - NLM
STAT- PubMed-not-MEDLINE
LR  - 20220915
IS  - 2169-3536 (Print)
IS  - 2169-3536 (Electronic)
IS  - 2169-3536 (Linking)
VI  - 10
DP  - 2022
TI  - Human Motion Enhancement via Tobit Kalman Filter-Assisted Autoencoder.
PG  - 29233-29251
LID - 10.1109/access.2022.3157605 [doi]
AB  - We present a novel approach to enhance the quality of human motion data collected 
      by low-cost depth sensors, namely D-Mocap, which suffers from low accuracy and 
      poor stability due to occlusion, interference, and algorithmic limitations. Our 
      approach takes advantage of a large set of high-quality and diverse Mocap data by 
      learning a general motion manifold via the convolutional autoencoder. In 
      addition, the Tobit Kalman filter (TKF) is used to capture the kinematics of each 
      body joint and handle censored measurement distribution. The TKF is incorporated 
      with the autoencoder via latent space optimization, maintaining adherence to the 
      motion manifold while preserving the kinematic nature of the original motion 
      data. Furthermore, due to the lack of an open source benchmark dataset for this 
      research, we have developed an extension of the Berkeley Multimodal Human Action 
      Database (MHAD) by generating D-Mocap data from RGB-D images. The newly extended 
      MHAD dataset is skeleton-matched and time-synced to the corresponding Mocap data 
      and is publicly available. Along with simulated D-Mocap data generated from the 
      CMU Mocap dataset and our self-collected D-Mocap dataset, the proposed algorithm 
      is thoroughly evaluated and compared with different settings. Experimental 
      results show that our approach can improve the accuracy of joint positions and 
      angles as well as skeletal bone lengths by over 50%.
FAU - Lannan, Nate
AU  - Lannan N
AUID- ORCID: 0000-0002-7255-0453
AD  - School of Electrical and Computer Engineering, Oklahoma State University, 
      Stillwater, OK 74078, USA.
FAU - Zhou, L E
AU  - Zhou LE
AD  - School of Electrical and Computer Engineering, Oklahoma State University, 
      Stillwater, OK 74078, USA.
FAU - Fan, Guoliang
AU  - Fan G
AUID- ORCID: 0000-0002-8584-9040
AD  - School of Electrical and Computer Engineering, Oklahoma State University, 
      Stillwater, OK 74078, USA.
LA  - eng
GR  - R15 AG061833/AG/NIA NIH HHS/United States
PT  - Journal Article
DEP - 20220308
PL  - United States
TA  - IEEE Access
JT  - IEEE access : practical innovations, open solutions
JID - 101639462
PMC - PMC9455937
MID - NIHMS1791233
OTO - NOTNLM
OT  - Autoencoder
OT  - Tobit Kalman filter
OT  - depth sensors
OT  - human motion manifold
OT  - motion capture
EDAT- 2022/09/13 06:00
MHDA- 2022/09/13 06:01
PMCR- 2022/09/08
CRDT- 2022/09/12 03:43
PHST- 2022/09/12 03:43 [entrez]
PHST- 2022/09/13 06:00 [pubmed]
PHST- 2022/09/13 06:01 [medline]
PHST- 2022/09/08 00:00 [pmc-release]
AID - 10.1109/access.2022.3157605 [doi]
PST - ppublish
SO  - IEEE Access. 2022;10:29233-29251. doi: 10.1109/access.2022.3157605. Epub 2022 Mar 
      8.