TY - GEN
T1 - From Image to Stability
T2 - 16th European Conference on Computer Vision, ECCV 2020
AU - Scott, Jesse
AU - Ravichandran, Bharadwaj
AU - Funk, Christopher
AU - Collins, Robert T.
AU - Liu, Yanxi
N1 - Funding Information:
Acknowledgments. We thank our collaborator Professor John Challis for offering his Kinesiology perspectives and helpful discussions. We thank Master Sitan Chen of Sitan Tai Chi School; Dr. Hesheng Bao of Win-Win Kung Fu Culture Center, Inc.; Professor Zuofeng Sun of School of Physical Education, Hebei Normal University; and Professor Pingping Xie of Tianjin University of Sport for their expertise and support of many years. We thank all the human subject volunteers for participating in our Taiji data collection. This human study is carried out under Penn State University IRB #8085 and was supported in part by NSF grant IIS-1218729 and the College of Engineering Dean’s office of Penn State University.
PY - 2020
Y1 - 2020
N2 - We propose and validate two end-to-end deep learning architectures to learn foot pressure distribution maps (dynamics) from 2D or 3D human pose (kinematics). The networks are trained using 1.36 million synchronized pose+pressure data pairs from 10 subjects performing multiple takes of a 5-min long choreographed Taiji sequence. Using leave-one-subject-out cross validation, we demonstrate reliable and repeatable foot pressure prediction, setting the first baseline for solving a non-obvious pose to pressure cross-modality mapping problem in computer vision. Furthermore, we compute and quantitatively validate Center of Pressure (CoP) and Base of Support (BoS), two key components for stability analysis, from the predicted foot pressure distributions.
AB - We propose and validate two end-to-end deep learning architectures to learn foot pressure distribution maps (dynamics) from 2D or 3D human pose (kinematics). The networks are trained using 1.36 million synchronized pose+pressure data pairs from 10 subjects performing multiple takes of a 5-min long choreographed Taiji sequence. Using leave-one-subject-out cross validation, we demonstrate reliable and repeatable foot pressure prediction, setting the first baseline for solving a non-obvious pose to pressure cross-modality mapping problem in computer vision. Furthermore, we compute and quantitatively validate Center of Pressure (CoP) and Base of Support (BoS), two key components for stability analysis, from the predicted foot pressure distributions.
UR - http://www.scopus.com/inward/record.url?scp=85097374521&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85097374521&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-58592-1_32
DO - 10.1007/978-3-030-58592-1_32
M3 - Conference contribution
AN - SCOPUS:85097374521
SN - 9783030585914
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 536
EP - 554
BT - Computer Vision – ECCV 2020 - 16th European Conference, Glasgow, 2020, Proceedings
A2 - Vedaldi, Andrea
A2 - Bischof, Horst
A2 - Brox, Thomas
A2 - Frahm, Jan-Michael
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 23 August 2020 through 28 August 2020
ER -