Logo

ITU AI/ML in 5G Challenge Webinar: Lightning-Fast Modulation Classification with Hardware-Efficient Neural Networks - Shared screen with speaker view
Jingyi Li
20:39
We will have recording?
Michaela Blott
20:53
yes, this will be recorded
Vishnu Ram
22:47
yes, this will be recorded and recording will be uploaded in the same webinar page.
Thomas Basikolo - ITU
24:12
Geneva, Switzerland
Ussama Zahid
24:22
Dublin, Irelan
Armand Kamary
24:22
Montreal, Canada
Khalid Albagami
24:24
Riyadh, Saudi Arabia
Rajesh Bansal
24:24
Singapore
Andrew Maclellan
24:24
Hi everyone. Glasgow, Scotland
Paolo Novellini
24:26
Milano, Italy
James Garland
24:27
Carlow, Ireland
Alessandro Pappalardo
24:27
Bergamo, Italy
Shreejith Shanker
24:29
Dublin, Ireland.
Biswadeb Dutta
24:30
Boston, USA
Miriam Leeser
24:32
Boston, MA
Michaela Blott
24:33
Dublin, Ireland
Huy T. Nguyen
24:34
Singapore
Kees Vissers
24:34
San Jose, CA, USA
Vishnu Ram
24:36
Bangalore, India
Felix Jentzsch
24:42
Bielefeld, Germany
Yaman Umuroglu
24:43
Oslo, Norway
Markus Maaß
24:59
Kisselbach, Germany
Rajesh Bansal
01:03:08
How did the 56% accuracy arrived for evaluation ?
Thomas Basikolo - ITU
01:03:48
https://github.com/Xilinx/brevitas-radioml-challenge-21
Thomas Basikolo - ITU
01:04:09
Thank you very much for joining today’s webinar.If you have a question, please type in the Q&A.Registration:The Challenge registration is open until 31 August, 2021: https://challenge.aiforgood.itu.int/Select problem statements: You can select problem statements here: https://challenge.aiforgood.itu.int/matchAlternatively : https://2ja3zj1n4vsz2sq9zh82y3wi-wpengine.netdna-ssl.com/wp-content/uploads/2021/06/Problem_statements_2021_v2.pdfChallenge Website: https://aiforgood.itu.int/ai-ml-in-5g-challenge/Today’s event page: https://aiforgood.itu.int/events/lightning-fast-modulation-classification-with-hardware-efficient-neural-networks/Next event/webinar: https://aiforgood.itu.int/events/radio-strike-a-reinforcement-learning-game-for-mimo-beam-selection-in-unreal-engine-3-d-environments/Join our Slack Channel to keep updatedhttps://join.slack.com/t/itu-challenge/shared_invite/zt-eql00z05-CXelo7_aL0nHGM7xDDvTmA
Thomas Basikolo - ITU
01:10:42
On the Challenge Platform, the problem statement can be accessed here: https://challenge.aiforgood.itu.int/match/matchitem/34
Yaman Umuroglu
01:26:59
also copy-pasting in case it gets lost in the Q&A section:
Yaman Umuroglu
01:27:00
Anonymous Attendee 04:39 PMI have a small question, is there a hard requirement to use brevitas and finn? E.g. if a special/some new layer is not available in pytorch, onnx and finn? Can we develop based on our own?This question has been answered liveYaman Umuroglu (You) 05:03 PMThere is a requirement to use Brevitas, but there is some room for custom operations -- please look under the Evaluation Criteria on the challenge webpage, under Custom Operations