Amortized Inverse Kinematics via Graph Attention for Real-Time Human Avatar Animation

2026-04-17

AI Summary

This work addresses the limitations of traditional inverse kinematics in real-time performance, robustness to input noise, and ambiguity in twist rotation by proposing IK-GAT, a lightweight graph attention network. IK-GAT recovers full joint orientations from sparse 3D keypoints in a single forward pass over a parent-child skeletal graph. The method explicitly models the twist axis using bone-aligned world-coordinate rotation representations, combined with continuous 6D rotation encoding, an SO(3) geodesic loss, and a forward kinematics consistency regularizer. With only 374K parameters, the model achieves over 650 FPS on CPU, significantly outperforming iterative approaches like VPoser without warm-start initialization, while demonstrating strong robustness to both input noise and initial pose variations.

Technology Category

• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Biometrics, Face, Gesture & Pose

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Inverse kinematics (IK) is a core operation in animation, robotics, and biomechanics: given Cartesian constraints, recover joint rotations under a known kinematic tree. In many real-time human avatar pipelines, the available signal per frame is a sparse set of tracked 3D joint positions, whereas animation systems require joint orientations to drive skinning. Recovering full orientations from positions is underconstrained, most notably because twist about bone axes is ambiguous, and classical IK solvers typically rely on iterative optimization that can be slow and sensitive to noisy inputs. We introduce IK-GAT, a lightweight graph-attention network that reconstructs full-body joint orientations from 3D joint positions in a single forward pass. The model performs message passing over the skeletal parent-child graph to exploit kinematic structure during rotation inference. To simplify learning, IK-GAT predicts rotations in a bone-aligned world-frame representation anchored to rest-pose bone frames. This parameterization makes the twist axis explicit and is exactly invertible to standard parent-relative local rotations given the kinematic tree and rest pose. The network uses a continuous 6D rotation representation and is trained with a geodesic loss on SO(3) together with an optional forward-kinematics consistency regularizer. IK-GAT produces animation-ready local rotations that can directly drive a rigged avatar or be converted to pose parameters of SMPL-like body models for real-time and online applications. With 374K parameters and over 650 FPS on CPU, IK-GAT outperforms VPoser-based per-frame iterative optimization without warm-start at significantly lower cost, and is robust to initial pose and input noise

Problem

• Inverse Kinematics
• Real-Time Animation
• Joint Orientation Estimation
• Human Avatar
• Underconstrained Problem

Innovation

• Amortized Inverse Kinematics
• Graph Attention Network
• Real-Time Avatar Animation
• Bone-Aligned Rotation Representation
• SO(3) Geodesic Loss