MILC 2018 - 2018 Microsoft Indoor Localization Competition

2018 Microsoft Indoor Localization Competition
*Registration Deadline: February 11th, 2018*
April 11-13, 2018
Porto, Portugal
Co-located with IPSN 2018
https://www.microsoft.com/en-us/research/event/mic...
===
ELIGIBILITY
Both academia and industry submissions are encouraged. All location
techniques, such as ranging, fingerprinting, infrastructure, or device
free, are welcome, except those that require end users’ manual
measurements. Contesters can deploy their own infrastructure of up to
10 devices (this number is tentative and it will be finalized after
the submission deadline). Normal RF interference is expected, but no
jammers from other deployments are allowed. The results must be shown
on a portable device, such as a phone or a tablet/laptop that a person
can easily carry around.
Demo submissions that do not meet one or more of the guidelines above
will be included in the poster session and will be evaluated as a
regular submission, but they will not be considered for prizes.
The competition will take place if at least 5 teams respond to this
preliminary call for competition.
COMPETITION CATEGORIES
Depending on the nature and number of submissions multiple categories
might be defined based on the accuracy (i.e., point-based vs. area
based), the size, the cost, or the type (i.e., software vs. hardware)
of the proposed solution.
Given our past experience, this year we expect to have the following
categories (note that these categories have not been finalized. The
final categories will be announced shortly after the registration
deadline and they will depend on the number and type of submissions
received):
Commercial off-the-shelf (COTS) Technologies: Submissions in this
category should be able to work with unmodified commercial
off-the-shelf devices such as laptops, phones, and tablets. In this
category, the unmodified COTS device is localized. Teams in this
category will not be allowed to interface any custom hardware to the
COTS devices (i.e., UWB or ultrasound hardware). Only changes to the
software of the devices will be allowed. Submissions in this category
could be further classified to infrastructure-based and
infrastructure-free depending on their requirement to deploy custom
hardware (i.e., BLE beacons) in the evaluation area.
Commercial off-the-shelf (COTS) Technologies with Initialization:
Submissions in this category should meet all the criteria of the COTS
Technologies category, with the additional requirement to initialize
the location of the COTS device being localized to a ground truth
location before the evaluation.
Modified Commercial off-the-shelf (COTS) Technologies: Submissions in
this category could interface custom hardware to the COTS devices to
be localized (i.e., UWB or ultrasound hardware). Submissions in this
category could be further classified to infrastructure-based and
infrastructure-free depending on their requirement to deploy custom
hardware (i.e., BLE beacons) in the evaluation area. Most, if not all,
submissions in this category are expected to be infrastructure-based.
For instance:
WiFi fingerprinting approaches that also leverage inertial sensors
will be classified as infrastructure-free COTS Technologies as they do
not require to interface any custom hardware to the COTS device being
localized.
WiFi fingerprinting approaches that also deploy custom BLE beacons in
the evaluation area will be classified as infrastructure-based COTS
Technologies as they don’t modify the COTS device being localized, but
they require the deployment of custom hardware.
Approaches based on inertial sensing that require to be initialized to
a ground truth location before being evaluated will be classified as
COTS Technologies with Initialization.
UWB approaches will be classified as infrastructure-based modified
COTS Technologies as they require modifications to the COTS device
being localized, and they need to deploy custom hardware in the
evaluation area.
In this year’s competition, teams will be required to report 3D
locations (X,Y,Z). The evaluation area we have reserved, even though
it does not contain multiple floors, it includes locations with
different elevation characteristics. On top of that, we plan to place
the devices to be localized at different heights during the
competition to evaluate the 3D localization accuracy of each team.
Not all teams will have to report 3D locations. We understand that
some technologies are not suited for 3D localization (i.e.,
WiFi/geo-magnetic fingerprinting, dead reckoning, BLE-based etc.), but
they are still excellent candidates for commercial indoor location
systems. With this in mind, the organizers will determine shortly
after the registration deadline which teams will be classified as 3D
localization teams. It is not up to the individual teams to decide if
they will report 2D or 3D locations. The organizers will ensure that
all the teams in a given category will either report 2D or 3D
locations. There will be no categories containing both 2D and 3D
localization teams.
ACCEPTABLE TECHNOLOGIES
LiDAR-based systems will not be accepted to the competition this year.
LiDAR is reserved for ground truth measurements. Also, the use of
laser range finders is allowed only during the setup of each system
and not during its evaluation.
Other than the LiDAR and laser range finder restrictions outlined
above, there are no other technology restrictions.
EVALUATION AND PRIZE
Results are judged based on accuracy, and an award will be given to
the top 3 teams in each category. When accuracy ties, infrastructure
requirements will be used for tie breaking. The winning teams in each
category will be invited to present their approach at the conference,
and receive a cash award. The exact accuracy metrics that will be used
during evaluation will be announced shortly before the competition
takes place.
Traditionally, teams were evaluated on their accuracy in estimating
locations of individual points. This year, teams will be evaluated on
their tracking accuracy as opposed to point accuracy. Each system
under test will be taken through a pre-determined path and will output
its estimated location at a predetermined rate while walking along the
path. The system that will be able to approximate the ground truth
path with the highest overall accuracy wins. Organizers will provide
more details about the evaluation process and its logistics as we get
closer to the competition.
POSTER SESSION
A poster session dedicated to all competition participants will be
organized during the conference. Participants will have the
opportunity to explain their system to conference attendees.
SUBMISSION GUIDELINES
Contesters must submit an abstract describing their approach and
deployment requirements by the contest registration deadline.
Submissions are treated as confidential until the competition.
Submissions must be at most two (2) single-spaced 8.5″ x 11″ pages,
including figures, tables, and references. Submission should follow
the exact same format as regular, full IPSN 2018 papers. Abstracts
should include the names and affiliations of all authors. Templates
can be found: http://ipsn.acm.org/2018/submission.html?v=1 Abstracts
should be sent over email to: dlymper-AT-microsoft.com on or before
February 11th 2018 with the following subject line: 2018 Microsoft
Indoor Localization Competition Submission.
REGISTRATION DEADLINE
February 11th, 2018, 11:59 pm EST
To register for the competition, please email your abstract to
dlymper-AT-microsoft.com with the following subject line: 2018 Microsoft
Indoor Localization Competition Submission.