SIGGRAPH Asia 2022 Conference Papers 2022
DOI: 10.1145/3550469.3555428
|View full text |Cite
|
Sign up to set email alerts
|

Transformer Inertial Poser: Real-time Human Motion Reconstruction from Sparse IMUs with Simultaneous Terrain Generation

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
22
0

Year Published

2023
2023
2024
2024

Publication Types

Select...
4
3
1

Relationship

1
7

Authors

Journals

citations
Cited by 51 publications
(23 citation statements)
references
References 40 publications
1
22
0
Order By: Relevance
“…Yi et al [40] took the same input, but generated both accurate pose and precise global translation. More recently, Jiang et al [15] not only accurately estimated the full-body motion but also handled the joint and global position drift that most IMU systems suffer from.…”
Section: Self-avatars and Animation Fidelitymentioning
confidence: 99%
“…Yi et al [40] took the same input, but generated both accurate pose and precise global translation. More recently, Jiang et al [15] not only accurately estimated the full-body motion but also handled the joint and global position drift that most IMU systems suffer from.…”
Section: Self-avatars and Animation Fidelitymentioning
confidence: 99%
“…Transformer-based models [Vaswani et al 2017], initially proposed for natural language processing, have been extensively used in many domains with sequential inputs. In that manner, Jiang et al [2022b] introduce a conditional Transformer decoder model that reconstructs full-body pose and can correct the drift by predicting stationary body points with soft-IK constraints, stabilizing the generated root velocity and joint angles. Apart from kinematic models, physics-based methods have also been used for motion reconstruction with IMUs.…”
Section: Full-body Motion Reconstruction From Sparse Inputmentioning
confidence: 99%
“…Other works [Huang et al 2018;Jiang et al 2022b;Yi et al 2022Yi et al , 2021 use a sparse set of IMUs (e.g., six) to reduce the cost of motion capture systems such as Xsens (17 IMUs) while still being able to represent a broader range of motion by placing some sensors on the lower body. IMU-based approaches have become increasingly popular due to their advantages in certain applications.…”
Section: Introductionmentioning
confidence: 99%
“…as Inertial Measurement Unit (IMU) devices, e.g., [Huang et al 2018;Jiang et al 2022;von Marcard et al 2017].…”
Section: Human Motion Trackingmentioning
confidence: 99%
“…As a human motion tracking device, these are handicapped by the lack of sensory information regarding the lower body and legs, which are essential to synthesizing believable full-body motion. Multiple methods have been proposed to address this, using transformers [Jiang et al 2022;Vaswani et al 2017], VAEs [Dittadi et al 2021] and normalizing flows generative models [Aliakbarian et al 2022]. Being kinematic-based approaches, however, these methods do not enforce physical properties and thus suffer from motion artifacts such as foot-skating and jitter.…”
Section: Human Motion Trackingmentioning
confidence: 99%