Art and anatomy based techniques for interactive character modeling and animation

角色建模和动画具有无限的可能性，在医疗、教育、培训、电信、人体工程学、人机交互以及最后但并非最不重要的娱乐领域具有重要应用。然而，目前的动画技术对于动画师、临床医生和一般用户来说是次优的，因为它们是由数字几何的数学和运动物理驱动的，而不是对解剖学、美学和用户的特定交互技能的理解。这个项目将探索交互式角色建模、动画和重复使用的技术，基于解剖学、美学、计时的基础，并利用他们特定的用户技能来建模和制作数字角色的动画。计算机呈现角色的最新技术就像一个提线木偶，允许动画师通过角色装备上类似木偶的控制来进行情感表达。操纵角色是一个复杂的过程，角色和面部操纵的解剖方法已经显示出在角色重用和捕捉动画中用其他技术难以捕捉的微妙之处方面的前景，这是本项目希望探索的。更大的挑战是控制一个被操纵的角色的动画，我们希望设计出由导演与演员互动并指定演员动作的高级方式驱动的技术。手的动作和手势的复杂性使其不仅是一个具有挑战性的动画问题，而且是人类功能、交流和审美的普遍组成部分。难以及早诊断的重复性劳损极大地损害了工作场所的工作效率。手的解剖学精确计算模型的开发是该项目的重要部分。解剖手部模型的最终目标是捕捉肌肉驱动部位(概念上相当于收缩冲动)的数值输入与手部骨骼关节运动之间的函数关系。映射将是正向的，其中冲量值可以被映射到结果的关节运动，以及反向，在那里我们计算各种肌肉组合中的许多可能的收缩脉冲，这些肌肉组合产生关节的一些给定运动。这种针对特定患者的映射可以是临床环境中有价值的运动分析工具，也可以是用于创建和验证逼真手部动画的工具。

Character modeling and animation has limitless possibilities with important applications in medicine, education, training, telecommunication, ergonomics, human computer interaction and last but not least entertainment. Current animation techniques, however, are suboptimal for animators, clinicians and users in general since they are driven by the mathematics of digital geometry and the physics of motion rather than an understanding of anatomy, aesthetics and the specific interactive skills of the users. This project will explore techniques for interactive character modeling, animation and reuse based on foundations of anatomy, aesthetics, timing and leveraging their specific user skillsets towards the modeling and animation of digital characters. The state of the art in the computer representation of a character resembles a marionette, allowing an animator to emote through puppet-like controls on the character rig. Rigging a character is a complex process and anatomic approaches to character and facial rigging have shown promise for character reuse and in capturing subtleties in animation that are difficult to capture with other techniques, that this project hopes to explore. The bigger challenge is controlling the animation of a rigged character, where we hope to devise techniques driven by the high level manner in which directors would interact with and specify the motion of actors. The complexity of motion and gestures of the hand makes it not only a challenging animation problem but a pervasive part of human function, communication and aesthetics. Repetitive strain injuries, which are hard to diagnose early, greatly impair productivity in the workplace. The development of an anatomically accurate computational model of the hand is an important part of this project. The end goal of the anatomic hand model is to capture the functional relationship between numerical inputs to the muscle actuation sites (conceptually equivalent to contraction impulses) and the motion of the skeletal joints of the hand. The mapping will be both forward, where impulse values can be mapped to a resulting joint motion, and inverse, where we compute the many possible contraction impulses in various muscle combinations that produce some given motion of the joints. Such a mapping for a specific patient can be a valuable motion analysis tool in a clinical setting as well as a tool for the creation and validation of realistic hand animation.