An SVM learning approach to robotic grasping

By Raphael Pelossof, Andrew T. Miller, Peter K. Allen, Tony Jebara. 2004

255 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Control and Systems Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Control and Systems Engineering

Citations in the corpus

Data-Driven Grasp Synthesis—A Survey (2013, methodology): “on this object part. Grasp experience is therefore only used to decide where to apply a grasp, not how the grasp should be conﬁgured. These steps are shown for two objects in Fig. 10. Pelossof et al. [20] approximate an object with a a single SQ. Given this, their goal is to ﬁnd a suitable grasp con”
VisGraB: A benchmark for vision-based grasping (2012, methodology): “, [6, 7, 12, 26]), whereas others are performed in the real world (e.”
Domain Randomization and Generative Models for Robotic Grasping (2017, background): “Most techniques estimate the value of the quality metric in real-world trials using traditional computer vision and learning techniques [21], [35], [43], [44].”
ContactGrasp: Functional Multi-finger Grasp Synthesis from Contact (2019, methodology): “We use the negated surface normal −n at the approach point a as the approach vector, making since the hand approach anti-parallel to the object surface normal (similar to [39]).”
Active learning using mean shift optimization for robot grasping (2009, background): “Pelossof et al. [13] uses an analytical model to determine local features of a good grasp, and subsequently interpolates to untested situations in simulation.”
Model learning for robot control: a survey (2011, background): “2007), grasping (Pelossof et al. 2004), robot control (Ferreira et al.”
The Columbia grasp database (2009, background): “In general, the relation between hand pose and grasp quality for a given object is both nonlinear and discontinuous, and more sophisticated learning methods such as SVMs have so far been shown to work only for simple objects [16], [2].”
Efficient grasping from RGBD images: Learning using a new rectangle representation (2011, methodology): “[16] used support vector machines to estimate the quality of a grasp given a number of features based on spin images; however this work was done in simulation and it is not clear how it would apply to real-world grasping.”
An overview of 3D object grasp synthesis algorithms (2012, methodology): “[61] used support vector machines to build a regression mapping between object shape, grasp parameters and grasp quality (Fig.”
Grasping unknown objects using an Early Cognitive Vision system for general scene understanding (2011, methodology): “A SVM has been trained to predict the grasp quality based on the hand configuration and the parameters of a single-superquadric representation of the objects in [16].”
Deep-learning in Mobile Robotics - from Perception to Control Systems: A Survey on Why and Why not (2016, methodology): “D model (Miller and Allen, 2004; Le´on et al., 2010; Pelossof et al., 2004), where a strong prior information is necessary.”
Hierarchical Fingertip Space: A Unified Framework for Grasp Planning and In-Hand Grasp Adaptation (2016, background): “More recent work demonstrates topological analysis of shape for grasping and caging [21], [22].”