[robotics-worldwide] [software] SLAMBench 2.0: a Benchmarking Framework for SLAM Algorithms

Sajad Saeedi s.saeedi at imperial.ac.uk
Fri Mar 16 09:08:37 PDT 2018


It is our pleasure to announce the release of the SLAMBench 2.0 
framework that allows you to compare different SLAM algorithms in 
astandardised way.

The framework can be used for different purposes, such as:
-*Benchmarking*: compare your algorithm with others'
-*Design Space Exploration (DSE)*: explore the effect of different 
parameters on the performance of the algorithm in an easy way
-*Build robotic systems*: You need SLAM to build an autonomous robot. 
SLAMBench 2 helps you to decide which algorithm works the best with your 
hardware.

Features of the framework:
- Standard API to work with any SLAM algorithm
- Easy to add new algorithms
- Easy to test multiple SLAM algorithms on one datasets
- Real-time performance
- Support for various dataset formats such as TUM-RGBD, EuRoC, ICL-NUIM, ...
- Currently these algorithms are supported (listed alphabetically):
--*ElasticFusion*
--*InfiniTAM*
--*KinectFusion*
--*LSDSLAM*
--*MonoSLAM*
--*OKVIS*
--*ORBSLAM2*
--*PTAM*
--*SVO1*

- You can easily measure several well-known performance metrics such as
--Absolute Trajectory Error (ATE)
-- Relative Pose Error (RPE)
-- Reconstruction Error (RER)
-- Power Consumption
-- Computation Speed
-- Memory Usage

See the paper here (to be published at ICRA 2018):
https://urldefense.proofpoint.com/v2/url?u=https-3A__www.research.ed.ac.uk_portal_files_54741939_main.pdf&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=P_7FMMqQmsThlkjV6binkR8SK2tw6x-_hUdt6b1bgUQ&e= 
<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.research.ed.ac.uk_portal_files_54741939_main.pdf&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=P_7FMMqQmsThlkjV6binkR8SK2tw6x-_hUdt6b1bgUQ&e=>

Software is available here:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_pamela-2Dproject_slambench2&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=D0vmqTOFUHePLJWMTL1se6I0CsrsrHs2ty6eiOLcYcw&e= 
<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_pamela-2Dproject_slambench2&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=D0vmqTOFUHePLJWMTL1se6I0CsrsrHs2ty6eiOLcYcw&e=>

See a 2-minute introduction on Youtube:
https://urldefense.proofpoint.com/v2/url?u=https-3A__youtu.be_-5FnDFVHWvg2Y&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=s-DLJfXfQ7BiGRqKtDi-pkCTjYukWgZASQ5yF1rVSd8&e=

SLAMBench 2.0 is a product of the PAMELA project, a joint project 
between the University of Edinburgh, Imperial College, and the 
University of Manchester.

-- 

Regards, Sajad Saeedi, PhD Research Associate Imperial College London 
London UK Website: https://urldefense.proofpoint.com/v2/url?u=http-3A__www.sajad-2Dsaeedi.ca_&d=DwIDaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=h-t1NdMcLxNzqGVgY5HoJmRTa6rIk_eqBgrIMUDimOM&s=8KAsoFRmdxR5o6z0FqUt33hjacCGnk5M9lhFNELyPwg&e=



More information about the robotics-worldwide mailing list