TOMSY
Topology Based Motion Synthesis for Dexterous Manipulation
Resources
- Grasp taxonomy Abstract: "The generation of a comprehensive grasp taxonomy", Feix, Pawlik, Schmiedmayer, Romero, Kragic
- Karlsruhe Object Database
- OpenGrasp Simulator
- Vision based tracking benchmark dataset
- EXtensible Optimization Toolset (EXOTica)
- Multi-modal high-quality manipulation data for benchmarking
- An executable program to guide wrapping movements. A system that demonstrates the wrapping procedure of deformable objects using the electric flux.
GraspLib
Portable library for rendering and recognizing articulations of human hand.
- Languages: C++, MATLAB
- Platforms: MacOS X, Ubuntu Linux
- Build system: CMake
- Dependencies: Boost, OpenCV, OGRE
Rigid Pose Estimation
The video above shows examples from the stereo benchmark dataset
corresponding to the paper:
Pauwels, Karl, Rubio, Leo, Diaz Alonso, Javier, Ros, Eduardo(2013):
Real-time model-based rigid object pose estimation and tracking
combining dense and sparse visual cues. In: Proceedings of the IEEE
Computer Society Conference on Computer Vision and Pattern Recognition
(CVPR), Portland, 2013.
The dataset consists of stereo video sequences featuring six different
objects (varying in shape and appearance) in scenarios involving
realistic object and camera motion, background clutter, image noise, and
occlusions. The dataset is available at
http://www.karlpauwels.com/datasets/ together with 3D models,
ground-truth, and evaluation code.
Coordinator
Danica Kragic
Centre for Autonomous Systems
Computational Vision and
Active Perception Lab
School of Computer Science and Communication
KTH
10044, Stockholm, Sweden
tel: +468 790 6729
fax: +468 7230302
