Radar Challenge
“Human Activity Classification with Radar”
Hosted by 2020 IET International Radar Conference
Chongqing, China
www.ietradar.org/
Dr Julien Le Kernec (University of Glasgow) Dr Francesco Fioranelli (Technical University Delft),
Dr Shufan Yang (University of Glasgow), Prof. Olivier Romain (University Cergy-Pontoise).
NOTES:
Dataset available at http://researchdata.gla.ac.uk/848/
doi:10.5525/gla.researchdata.848
Fioranelli, F. , Shah, S. A. , Li, H., Shrestha, A., Yang, S. and Le Kernec, J. (2019) Radar sensing for healthcare. Electronics Letters, 55(19), pp. 1022-1024. (doi:10.1049/el.2019.2378)
You’ll find the details for the submission of classification results at the link below:
https://humanactivitiyclassificationwithradar.grand-challenge.org/
Prize
The prize for the winner of the DEMO night will be provided by the IET Technical network on Electromagnetics. This will be a book(s) from the IET library.
Contact
The challenge is organised by the 2020 IET International Radar Conference.
For enquiries regarding the challenge please contact:
Dr. Julien Le Kernec (julien.lekernec@glasgow.ac.uk)
Announcing the radar challenge
The Radar Challenge is a new event hosted at 2020 IET International Radar Conference that enables participants to test their classification algorithms on a common, publicly available database of radar data in order to benchmark performances. The participants will submit a regular conference paper for the Demo night describing their methodology and results and submit their code and performance criteria for the leader board. The challenge is based on an open dataset of raw radar data. The details of the radar setup and parameters are available with the dataset (see address below). You will have to pre-process the data and classify the data for which labels are provided. A training, validation, and test dataset will be defined for the competition. The goal is to build a community of classification algorithms developers for radar applications on the wider theme of “human activities”. The submissions for pre-selection will be submitted for inclusion in a special poster session at the conference. The 8 best entries according to the criteria below will be selected for the DEMO night, feel free to include performance criteria, you feel relevant in your paper submission.
The DEMO night
Up to 8 submissions will be selected to participate in the DEMO night at 2020 IET International Radar Conference. The participants will need to demonstrate live their classification algorithms on an unseen test dataset of raw radar data provided during the DEMO night. The participants will need to bring a poster and their platform for the DEMO. It is up to the participants to choose how they will showcase their work. Each will be given a time slot to explain their algorithms and demonstrate its performances. The participants will be judged on the performances of the algorithms but also the quality of the poster and demonstration. The winner at the DEMO night will be awarded a prize at the gala dinner.
Important dates
Challenge submission for the leader board from 01/02/2020 – 30/06/2020 (details of submission format as specified below)
Paper submission deadline: 30/06/2020 (all submission will be considered for inclusion in conference – 8 will be selected for the DEMO night)
NOTE: to be included in the DEMO night you need a submission both in the leader board and a paper submission
Paper acceptance notification: 31/07/2020
DEMO night acceptance notification: 31/07/2020
Camera-ready paper submission: 31/08/2020
new dataset release for DEMO night (01/10/2020):
Conference date: 4-6/11/2020 (DEMO night)
