Title:
METHOD FOR TRAINING SMPL PARAMETER PREDICTION MODEL, COMPUTER DEVICE, AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2020/156148
Kind Code:
A1
Abstract:
A method for training an SMPL parameter prediction model, comprising: obtaining a sample picture; inputting the sample picture to a posture parameter prediction model to obtain a posture prediction parameter; inputting the sample picture to a morphology parameter prediction model to obtain a morphology prediction parameter; computing a model prediction loss according to an SMPL prediction parameter and annotation information of the sample picture; and inversely training the posture parameter prediction model and the morphology parameter prediction model according to the model prediction loss.
Inventors:
SUN SHUANG (CN)
LI CHEN (CN)
TAI YU WING (CN)
JIA JIAYA (CN)
SHEN XIAOYONG (CN)
LI CHEN (CN)
TAI YU WING (CN)
JIA JIAYA (CN)
SHEN XIAOYONG (CN)
Application Number:
PCT/CN2020/072023
Publication Date:
August 06, 2020
Filing Date:
January 14, 2020
Export Citation:
Assignee:
TENCENT TECH SHENZHEN CO LTD (CN)
International Classes:
G06T13/40
Foreign References:
CN109859296A | 2019-06-07 | |||
CN109285215A | 2019-01-29 | |||
CN108594997A | 2018-09-28 | |||
CN201910103414A | 2019-02-01 |
Other References:
PENGFEI DOU, SHISHIR K. SHAH, AND IOANNIS A. KAKADIARIS: "End-to-end 3D Face Reconstruction with Deep Neural Networks", 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION, 9 November 2017 (2017-11-09), pages 1503 - 1512, XP055724139, DOI: 10.1109/CVPR.2017.164
See also references of EP 3920146A4
See also references of EP 3920146A4
Attorney, Agent or Firm:
ADVANCE CHINA IP LAW OFFICE (CN)
Download PDF: