Fast Grasp Synthesis for Various Shaped Objects

Abstract

Human-like grasp planning is difficult because a human hand has a high number of degrees of freedom, and there are many grasping styles depending on the shape of an object and purpose. We propose a fast grasp synthesis system which enables a user to choose the desired grasping styles from a set of grasp types in a human grasp taxonomy. Given a 3D model of an object, our system detects graspable positions and generates grasping hand postures in every applicable grasp types in the grasp taxonomy for each grasping position. Hand postures are generated separately for each digit, and hand alignment is then refined iteratively. A user can also specify the grasping position by moving the cursor onto object surface, as well as grasp type. The generated hand postures are shown as a table of thumbnail images, and the user can select the grasping hand posture by clicking on one of those postures. Our system enables interactive generation of various grasping hand postures in real time.

Physics‐based assistive grasping for robust object manipulation in virtual reality

“…However, the system is not for real‐time interactions; also, this work does not include any rich dimensional variations of the hand models . Kyota et al combined prerecorded grasp poses and grasp taxonomy for interactive grasp synthesis. Many methods use supervised or unsupervised learning techniques– to predict the stability of a grasp, but the shortcoming of these methods is that they require a sufficiently huge set of coherent labeled data …”

Robust realtime physics-based motion control for human grasping

“…An ideal grasping action must take into account the geometry and dynamic characteristics of the object to be grasped and the selection of contact between the object and the fingers, thumb, and palm of the hand. Recent years have seen some significant advances in this area [Pollard and Zordan 2005;Kry and Pai 2006;Li et al 2007;Liu 2009;Kyota and Saito 2012]. Still, the ultimate goal of building an automated realtime system that is capable of grasping a wide variety of objects with different geometry and physical quantities remains unsolved.…”

“…An appealing alternative to grasp synthesis is to use prerecorded grasp data [Elkoura and Singh 2003;Li et al 2007;Amor et al 2008;Kyota and Saito 2012]. This approach is appealing because synthesized grasp poses are often natural-looking and consistent with real world data.…”

“…Amor and colleagues [2008] constructed a probabilistic model for human grasp poses from prerecorded data and used it to constrain the solution space of grasp synthesis. Kyota and Saito [2012] combined prerecorded grasp poses and grasp taxonomy for interactive grasp synthesis. In addition, researchers have also explored how to use prerecorded motion data to synthesize finger motions for gesturing characters.…”

“…Our goal, however, is different because we aim to synthesize an entire sequence of human grasping action, including reaching, closing, and manipulation, while previous approaches are focused on synthesis of either single grasp poses [Elkoura and Singh 2003;Li et al 2007;Amor et al 2008;Kyota and Saito 2012] or freehand gestures [Jörg et al 2012]. This motivates us to develop a new datadriven model that compactly represents spatial-temporal behavior of human grasping.…”