Learning body shape variation in physics-based characters

Won, Jungdam; Lee, Jehee

doi:10.1145/3355089.3356499

Cited by 75 publications

(48 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Reverse curriculum learning has been shown to be effective at balancing uneven data generation in DRL. For example, [WL19] and [PRL*19] propose a form of adaptive sampling where more difficult tasks are given higher priority during training.…”

Section: Related Workmentioning

confidence: 99%

“…In this section, we provide the background for actor‐critic‐based policy‐gradient algorithms. Importantly, the critic module can be used to estimate the performance of the policy, as shown in [WL19, PRL* 19]. Our adaptive curriculum (§ 6.5) uses the critic to adjust the task difficulty.…”

Section: Learning Control Policiesmentioning

confidence: 99%

“…This strategy is equivalent to the adaptive sampling introduced in [PRL*19] and [WL19]. The idea is that during task sampling, more difficult tasks will cause more failure, leading to more frequent early termination.…”

Section: Learning Curriculamentioning

confidence: 99%

See 2 more Smart Citations

ALLSTEPS: Curriculum‐driven Learning of Stepping Stone Skills

Xie

Ling

Kim

et al. 2020

Computer Graphics Forum

View full text Add to dashboard Cite

Humans are highly adept at walking in environments with foot placement constraints, including stepping‐stone scenarios where footstep locations are fully constrained. Finding good solutions to stepping‐stone locomotion is a longstanding and fundamental challenge for animation and robotics. We present fully learned solutions to this difficult problem using reinforcement learning. We demonstrate the importance of a curriculum for efficient learning and evaluate four possible curriculum choices compared to a non‐curriculum baseline. Results are presented for a simulated humanoid, a realistic bipedal robot simulation and a monster character, in each case producing robust, plausible motions for challenging stepping stone sequences and terrains.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Learning Control Policiesmentioning

confidence: 99%

See 1 more Smart Citation

ALLSTEPS: Curriculum‐driven Learning of Stepping Stone Skills

Xie

Ling

Kim

et al. 2020

Computer Graphics Forum

View full text Add to dashboard Cite

show abstract

“…Recent works have explored deep learning methods for motion retargeting, such as recurrent neural networks [47] or deep reinforcement learning [49]. An important number of these works focus on video based motion retargeting.…”

Section: Data Driven Approachesmentioning

confidence: 99%

Contact preserving shape transfer: Retargeting motion from one shape to another

Basset

Wuhrer

Boyer

et al. 2020

Computers & Graphics

View full text Add to dashboard Cite

We present an automatic method that allows to retarget poses from a source to a target character by transferring the shape of the target character onto the desired pose of the source character. By considering shape instead of pose transfer our method allows to better preserve the contextual meaning of the source pose, typically contacts between body parts, than pose-based strategies. To this end, we propose an optimization-based method to deform the source shape in the desired pose using three main energy functions: similarity to the target shape, body part volume preservation, and collision management to preserve existing contacts and prevent penetrations. The results show that our method allows to retarget complex poses with several contacts to different morphologies, and is even able to create new contacts when morphology changes require them, such as increases in body size. To demonstrate the robustness of our approach to different types of shapes, we successfully apply it to basic and dressed human characters as well as wild animal models, without the need to adjust parameters. 1. Introduction 1 Animation studios have stored terabytes of animation files 2 applied on various 3D characters with meticulous association 3 between skeletal motion and 3D shape, created manually by 4 skilled artists who designed rigged models and corresponding 5 skeletal motions. Retargeting these existing skeletal motions 6 to new characters automatically is a long-standing problem in 7 computer animation [20] that remains challenging, especially 8 when the animation contains close interactions and contacts.

show abstract

“…Learning a control policy for a large set of unstructured motion capture data can be facilitated by motion matching [8] or RNN [7] based motion generators. In order to train a policy for more challenging tasks, adaptive sampling [7], [24] or curriculum learning [10], [12] in task or environmental parameter spaces have been proposed. Without using motion data, a controller learned by DRL has often exhibited a non-human-like gait [9].…”

Section: Related Workmentioning

confidence: 99%

Finite State Machine-Based Motion-Free Learning of Biped Walking

Kang

Lee

2021

IEEE Access

View full text Add to dashboard Cite

Recently, deep reinforcement learning (DRL) is commonly used to create controllers for physically simulated characters. Among DRL-based approaches, imitation learning for character control using motion capture clips as tracking references has shown successful results in controlling various motor skills with natural movement. However, the output motion tends to be constrained close to the reference motion, and thus the learning of various styles of motion requires many motion clips. In this paper, we present a DRL method for learning a finite state machine (FSM) based policy in a motion-free manner (without the use of any motion data), which controls a simulated character to produce a gait as specified by the desired gait parameters. The control policy learns to output the target pose for each FSM state and transition timing between states, based on the character state at the beginning of each step and the userspecified gait parameters, such as the desired step length or maximum swing foot height. The combination of FSM-based policy learning and simple linear balance feedback embedded in the base controller has a positive synergistic effect on the performance of the learned policy. The learned policy allows the simulated character to walk as instructed by the continuously changing the gait parameters while responding to external perturbations. We demonstrate the effectiveness of our approach through interactive control, external push, comparison, and ablation studies.

show abstract

Learning body shape variation in physics-based characters

Cited by 75 publications

References 29 publications

ALLSTEPS: Curriculum‐driven Learning of Stepping Stone Skills

ALLSTEPS: Curriculum‐driven Learning of Stepping Stone Skills

Contact preserving shape transfer: Retargeting motion from one shape to another

Finite State Machine-Based Motion-Free Learning of Biped Walking

Contact Info

Product

Resources

About