Data-driven glove calibration for hand motion capture

Abstract

Hand motion is an important component of human motion, playing a central role in communication. However, it is difficult to capture hand motion optically, especially in conjunction with full body motion. Due to a lack of appropriate calibration methods, data gloves also do not provide sufficiently accurate hand motion. In this paper, we present a novel glove calibration approach that can map raw sensor readings to hand motion data with both accurate joint rotations and fingertip positions. Our method elegantly handles the sensor coupling problem by treating calibration as a flexible mapping from sensor readings to joint rotations. A sampling process collects data tuples according to accuracy requirements, and organizes all the tuples in a training set. From these data, a specially designed Gaussian Process Regression model is trained to infer the calibration function, and the learned model can be used to calibrate new sensor readings. For real-time hand motion capture, a sparse approximation of the model is used to enhance performance. Evaluation experiments demonstrate that our approach provides significantly better results that have more accurate hand shapes and fingertip positions, compared to other calibration methods.

Assessing the Impact of Hand Motion on Virtual Character Personality

“…Hand animation for virtual characters can be reconstructed from image based [Zhao et al 2012;, marker based [Kang et al 2012;Wheatland et al 2013], and glove based [Wang and Neff 2013;Huenerfauth and Lu 2010;Griffin et al 2000;Hu et al 2004;Menon et al 2003;Steffen et al 2011] motion capture techniques, or synthesized using physics knowledge [Zhao et al 2013;Liu 2008;2009;Ye and Liu 2012], data-driven [Jörg et al 2012;ElKoura and Singh 2003] or rule-based methods [Zhu et al 2012]. However, despite all the hand motion acquisition and generation research, little hand motion perception work has been conducted.…”

“…[Samadani et al 2011] examined human performance in recognizing affective expressions of hand-like structures, and their follow-up work [Samadani et al 2013] developed a computational model for generating affective hand movements on anthropomorphic and non-anthropomorphic structures. In [Kessler et al 1995;Hoyet et al 2012], perceptual studies mainly examine the fidelity of the hand poses and motions, while [Wang and Neff 2013] added fingertip accuracy evaluation into their survey, but provide no answer to the question of which hand motions are best suited to a character's personality. This motivates our experiments on the impact of variations in hand poses and motions on the perception of personality.…”

“…The captured hand motion was mapped to a kinematic hand model which was previously proposed by [Kahlesz et al 2004;Griffin et al 2000;Hu et al 2004;Turner 2001;Wang and Neff 2013], to minimize motion loss and include more flexible thumb pronation. The kinematic model contains 23 DOFs (Figure 1), which generally matches the real human hand.…”

“…We ensure that the link length in the kinematic model is consistent with the size of the hand that performed the motion capture to reduce the retargetting error. Captured hand motions were calibrated through the method of [Wang and Neff 2013] All clips were rendered from the same, single camera viewpoint with the camera aimed at the palm of the hand. This was selected to ensure that the hand pose and movement could be clearly seen in the videos.…”

Automatic Hand-Over Animation using Principle Component Analysis

“…In particular, optical motion capture systems, while being very accurate, can require substantial post-processing to handle occlusions and mislabelings. Cyber Gloves [2013] and the like are robust to captures in larger spaces, but they require regular calibration [Wang and Neff 2013] and do not provide high enough accuracy for many applications [Kahlesz et al 2004]. For other camera-based or range-scan type systems, the hand needs to be in a confined space and the body can not captured synchronously [Wang and Popović 2009;Zhao et al 2012;].…”