Multi-Segment Foot Modeling for Human Animation

Abstract

We present a multi-segment foot model and its control method for the simulation of realistic bipedal behaviors. The ground reaction force is the only source of control for a biped that stands and walks on its feet. The foot is the body part that interacts with the ground and produces appropriate actuation to the body. Foot anatomy features 26 bones and many more muscles that play an important role in weight transmission, balancing posture and assisting ambulation. Previously, the foot model was often simplified into one or two rigid bodies connected by a revolute joint. We propose a new foot model consisting of multiple segments to accurately reproduce human foot shape and its functionality. Based on the new model, we developed a foot pose controller that can reproduce foot postures that are generally not obtained in motion capture data. We demonstrate the validity of our foot model and the effectiveness of our foot controller with a variety of foot motions in a physics-based simulation.

Humanoids’ Feet: State-of-the-Art & Future Directions

“…Multiple-segments feet model : models allow to calculate the kinematic parameters of the foot during walking [57,91,92]. Park et al [92] developed a five segments foot model with four 3 DoFs joints between the segments.…”

Functionality-Driven Musculature Retargeting

“…Arnold et al [2] and Holzbaur et al [14] provided a comprehensive reference of muscle modeling parameters for lower and upper limbs. Anatomical modeling for specific body parts, such as face [44,15], feet [37], hands [48,41], shoulders [50,18,33], tongue [47], and jaw [55], has been explored for last two decades. Comprehensive full-body musculoskeletal modeling and simulation systems are also available freely [7] and commercially [5].…”

From Universal Humanoid Control to Automatic Physically Valid Character Creation

“…Specifically, we first incorporate the human body shape into the state space of our controller and learn from a large-scale human motion dataset [Mahmood et al 2019] accompanied by realistic human body shape annotation. Based on the key observation that foot geometry and compliance are key contributors to agent instability [Park et al 2018] (29 muscles associated with the soft and compliable human foot vs 2 pieces of simple rigid bodies in most of the simulated characters), we propose a simple yet effective way to compensate for such a mismatch. We argue that while the multi-joint foot proposed in [Park et al 2018] is anatomically correct, humans wear shoes and footwear that are difficult to perfectly simulate.…”

“…Based on the key observation that foot geometry and compliance are key contributors to agent instability [Park et al 2018] (29 muscles associated with the soft and compliable human foot vs 2 pieces of simple rigid bodies in most of the simulated characters), we propose a simple yet effective way to compensate for such a mismatch. We argue that while the multi-joint foot proposed in [Park et al 2018] is anatomically correct, humans wear shoes and footwear that are difficult to perfectly simulate. Instead of adding additional joints to model a flexible foot, we apply residual external forces Yuan and Kitani 2020;Yuan et al 2021b] on the foot.…”

“…• is the elementwise multiplication. Based on the observation that foot compliance and flexibility are crucial in helping the humanoid stay stable[Park et al 2018], we resort to applying residual force on the foot geometry to best match the ground reacting force induced by foot muscle and shoes. For each foot geometry (ankle and toe as defined in SMPL), we apply five external forces 𝒆 𝑡 ≜ (𝒆 𝐴 𝑇 1𝑡 is the first external force applied to the ankle/toe at timestep t. For each external force 𝒆 𝐴 1𝑡 , it consist of the contact point position on the geom's local coordinate frame 𝒑 𝐴 1𝑡 ∈ R 3 , its direction 𝒅 𝐴…”