Generative GaitNet

Moon Seok Park2
Seoul National University1
Seoul National Univ. Bundang Hospital2

Responsive image

Interactive user interface for real-time gait simulation.


Understanding the relation between anatomy and gait is key to successful predictive gait simulation. In this paper, we present Generative GaitNet, which is a novel network architecture based on deep reinforcement learning for controlling a comprehensive, fullbody, musculoskeletal model with 304 Hill-type musculotendons. The Generative GaitNet is a pre-trained, integrated system of artificial neural networks learned in a 618-dimensional continuous domain of anatomy conditions (e.g., mass distribution, body proportion, bone deformity, and muscle deficits) and gait conditions (e.g., stride and cadence). The pre-trained GaitNet takes anatomy and gait conditions as input and generates a series of gait cycles appropriate to the conditions through physics-based simulation. We will demonstrate the efficacy and expressive power of Generative GaitNet to generate a variety of healthy and pathological human gaits in real-time physics-based simulation.


Jungnam Park, Sehee Min, Phil Sik Chang, Jaedong Lee, and Jehee Lee. 2022.
Generative GaitNet.
SIGGRAPH '22: ACM SIGGRAPH 2022 Conference Proceedings. Article No.: 22. (SIGGRAPH 2022)




                author = {Park, Jungnam and Min, Sehee and Chang, Phil Sik and Lee, Jaedong and Park, Moon Seok and Lee, Jehee},
                title = {Generative GaitNet},
                year = {2022},
                booktitle = {ACM SIGGRAPH 2022 Conference Proceedings},
                articleno = {22},
                series = {SIGGRAPH '22}