标题： 实时逆运动学用于生成虚拟人类角色的多约束运动 显示英文标题

标题： Real-Time Inverse Kinematics for Generating Multi-Constrained Movements of Virtual Human Characters

摘要： 在计算机图形学、交互式虚拟环境、机器人技术和生物力学的各种应用中，实时生成准确且逼真的虚拟人类运动具有重要意义。本文介绍了一种专为生成逼真类人运动而设计的新颖实时逆向运动学（IK）求解器。利用TensorFlow的自动微分和即时编译功能，所提出的求解器能够高效处理具有高自由度的复杂人体骨骼结构。通过将正向和逆向运动学视为可微操作，我们的方法有效解决了多约束问题中的常见挑战，如误差累积和复杂的关节限制，这些对于逼真的人体运动建模至关重要。我们在SMPLX人体骨骼模型上展示了求解器的有效性，将其性能与广泛使用的基于迭代的IK算法（如循环坐标下降（CCD）、FABRIK和非线性优化算法IPOPT）进行了比较。我们的实验涵盖了简单的末端执行器任务和具有真实关节限制的复杂多约束问题。结果表明，我们的IK求解器实现了实时性能，表现出快速收敛、每次迭代的计算开销最小，并且相比现有方法成功率有所提高。该项目代码可在https://github.com/hvoss-techfak/TF-JAX-IK获取。 显示英文摘要

摘要： Generating accurate and realistic virtual human movements in real-time is of high importance for a variety of applications in computer graphics, interactive virtual environments, robotics, and biomechanics. This paper introduces a novel real-time inverse kinematics (IK) solver specifically designed for realistic human-like movement generation. Leveraging the automatic differentiation and just-in-time compilation of TensorFlow, the proposed solver efficiently handles complex articulated human skeletons with high degrees of freedom. By treating forward and inverse kinematics as differentiable operations, our method effectively addresses common challenges such as error accumulation and complicated joint limits in multi-constrained problems, which are critical for realistic human motion modeling. We demonstrate the solver's effectiveness on the SMPLX human skeleton model, evaluating its performance against widely used iterative-based IK algorithms, like Cyclic Coordinate Descent (CCD), FABRIK, and the nonlinear optimization algorithm IPOPT. Our experiments cover both simple end-effector tasks and sophisticated, multi-constrained problems with realistic joint limits. Results indicate that our IK solver achieves real-time performance, exhibiting rapid convergence, minimal computational overhead per iteration, and improved success rates compared to existing methods. The project code is available at https://github.com/hvoss-techfak/TF-JAX-IK