[evaal contest] Some questions about the evaluation
이소연
sylee at etri.re.kr
Thu Jul 13 11:11:34 CEST 2017
Dear Estefania,
I am the chair of track 2, and please find the answers for your questions.
>I would like to know how do you need the output file: is it the whole track given as [timestamp, posX, posY, posZ] or this information but referred only to some points that you already have the ground truth position?
In track 2, we recommend the output file consisting of the whole track data with the format defined in the attached document(See IV. Description of Output File (Logfile)). Usually, sensor data sampling frequency can be 100 ~ 200Hz, but we recommend 50~100Hz as writing frequency for logfile.
>In the second case, we would need to implement some code to write down the markers time and position. Kind of pressing a button every time we walk on the markers, in a way that as output you only have the selected points.
Anyhow, the code to write down the index of markers is needed.
>We assume we start at (0,0,0) and start walking on the x direction, right?
In this competition, we use WGS84 coordinates. And, the coordinates of initial point is given to competitors. Even though you use relative coordinates inside an algorithm, you should translate it into absolute coordinates in WGS84 format.
>Is there loop closure in your defined trajectory? I mean, partial or total repetition of the trajectory.
Last year, the test track was formed with loop closure. Usually, the test track is finalized just before the competition, so I am not sure at this time. But, the building is relatively small size compared to last year, so, it is highly probable to include partial repetition of the trajectory.
>Only medium- and low-cost IMUs are allowed, right?
There is no constraint on the quality of IMU sensors. You can use any kinds of COTS product for the competition.
>How much time do the teams have to generate the output file after finishing the experiment? Are more than one repetitions allowed? >The final data will be transmitted via usb stick?
The output file was delievered to track chairs right after the experiment. If something is wrong during the experiment, then the second chance is given to the competitior. The final data is delivered via usb stick.
If you have any further questions, then please send me an E-mail.
Best regards,
Soyeon Lee
----------------------------------------------------------------------------------------------
SOYEON LEE (Principal Researcher/Ph.D)
Electronics and Telecommunications Research Institute (ETRI)
Intelligent Robotics Research Division
HMI Research Group
Tel: 042-860-1846 E-mail : sylee at etri.re.kr<mailto:sylee at etri.re.kr>
------------------------------------------------------------------------------------------------
________________________________
보낸 사람 : "Estefania.Munoz at dlr.de" <Estefania.Munoz at dlr.de>
보낸 날짜 : 2017-07-11 02:05:02 ( +09:00 )
받는 사람 : filippo.palumbo at isti.cnr.it <filippo.palumbo at isti.cnr.it>
참조 : contest at evaal.aaloa.org <contest at evaal.aaloa.org>
제목 : Re: [evaal contest] Some questions about the evaluation
Dear Filippo,
I was thinking of participating on Track 2.
Best regards, Estefanía.
From: Filippo Palumbo [mailto:filippo.palumbo at isti.cnr.it]
Sent: Monday, July 10, 2017 5:52 PM
To: Munoz Diaz-Ropero, Estefania
Cc: contest at evaal.aaloa.org
Subject: Re: [evaal contest] Some questions about the evaluation
Dear Estefanía,
I am the chair of track 1, do you need this information regarding track 1 ("Smartphone-based") or track 2 ("Pedestrian dead reckoning")?
Thank you for your interest in the IPIN 2017 competition, my best regards,
Filippo
On Mon, Jul 10, 2017 at 10:22 AM, <Estefania.Munoz at dlr.de<mailto:Estefania.Munoz at dlr.de>> wrote:
Dear evaluation team,
I would like to know how do you need the output file: is it the whole track given as [timestamp, posX, posY, posZ] or this information but referred only to some points that you already have the ground truth position?
In the second case, we would need to implement some code to write down the markers time and position. Kind of pressing a button every time we walk on the markers, in a way that as output you only have the selected points.
We assume we start at (0,0,0) and start walking on the x direction, right?
Is there loop closure in your defined trajectory? I mean, partial or total repetition of the trajectory.
Only medium- and low-cost IMUs are allowed, right?
How much time do the teams have to generate the output file after finishing the experiment?
Are more than one repetitions allowed?
The final data will be transmitted via usb stick?
Many thanks in advance for your time.
Best regards, Estefanía.
——————————————————————————
Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR)
German Aerospace Center
Institute of Communications and Navigation | Oberpfaffenhofen | 82234 Wessling | Germany
Dr. Estefanía Muñoz Díaz | Vehicular Applications
Telephone +49-8153-28-4132<tel:+49%208153%20284132> | Telefax +49-8153-28-1871<tel:+49%208153%20281871> | Estefania.Munoz at DLR.de<mailto:Estefania.Munoz at DLR.de>
www.DLR.de<http://www.dlr.de/>
_______________________________________________
Contest mailing list
Contest at evaal.aaloa.org<mailto:Contest at evaal.aaloa.org>
http://evaal.aaloa.org/mailman/listinfo/contest
clear="all">
--
Filippo Palumbo (Research Staff) Voice: +39.050.621.2958
ISTI - Area della ricerca CNR Mobile: +39.329.4477.405
via G. Moruzzi 1, I-56124 Pisa Skype: palumbo.f
(entrance 20, 1st floor, room C62) Web: http://www.isti.cnr.it/about/people.php?id=443<http://www.isti.cnr.it/about/people.php?id=443>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://evaal.aaloa.org/pipermail/contest/attachments/20170713/aefea087/attachment-0001.html>
More information about the Contest
mailing list