|Short Description:||Building a Benchmarking Toolkit for robotics|
|Collaborator:||GiulioFontana (firstname.lastname@example.org), DavideRizzi ()|
Rawseeds, or Robotics Advancement through Web-publishing of Sensorial and Elaborated Extensive Data Sets, has been funded by the European Union's 6th Framework Programme. Rawseeds built and published a high-quality Benchmarking Toolkit to compare and assess algorithms for robotics, mainly focused towards the problems of localization, mapping and SLAM (Simultaneous Localization And Mapping).
Rawseeds' Benchmarking Toolkit includes:
- high-quality data sets, obtained by exploring indoor and outdoor environments with a suitably equipped mobile robot;
- problems (called Benchmark Problems or BPs) defined on the above datasets, each of which includes methodologies to evaluate the performance of any algorithm used to solve the problem;
- algorithms (called Benchmark Solutions or BSs) that solve the BPs, along with their output when applied to the associated problems and with the results of the evaluation of such output using the methodologies defined in the BP.
See the official website of the project for details and downloads.
The Benchmarking Toolkit (500GB of compressed data!) has been completed and is online, but the work of Rawseeds is ongoing for what concerns the administration of the website and the management of the community of users.
Other AIRWiki users who worked on this project
The Rawseeds project was coordinated by AIRLab/Politecnico di Milano, and included three other partners:
Albert-Ludwigs-Universität Freiburg, Germany (Institut für Informatik)
Universidad de Zaragoza, Spain (Depto. Informática e Ingeniería de Sistemas)
Università degli Studi di Milano-Bicocca, Italy (I.R.A. Laboratory)