Machine learning competitions have grown in popularity and impact over the last decade, emerging as an effective means to advance the state of the art by posing well-structured, relevant, and challenging problems to the community at large. Motivated by a reward or merely the satisfaction of seeing their machine learning algorithm reach the top of a leaderboard, practitioners innovate, improve, and tune their approach before evaluating on a held-out dataset or environment. The competition track of NeurIPS has matured in 2020, its fourth year, with a considerable increase in both the number of challenges and the diversity of domains and topics. A total of 16 competitions are featured this year as part of the track, with 8 competitions associated to each of the two days. The list of competitions that are part of the program are available here: https://neurips.cc/Conferences/2020/CompetitionTrack
Competition Track Friday
Competition Track Saturday
Competition Track Friday
Hugo Jair Escalante, Katja Hofmann
Fri, Dec 11th @ 16:00 GMT – Sat, Dec 12th @ 01:45 GMT
Abstract: First session for the competition program at NeurIPS2020.
Machine learning competitions have grown in popularity and impact over the last decade, emerging as an effective means to advance the state of the art by posing well-structured, relevant, and challenging problems to the community at large. Motivated by a reward or merely the satisfaction of seeing their machine learning algorithm reach the top of a leaderboard, practitioners innovate, improve, and tune their approach before evaluating on a held-out dataset or environment. The competition track of NeurIPS has matured in 2020, its fourth year, with a considerable increase in both the number of challenges and the diversity of domains and topics. A total of 16 competitions are featured this year as part of the track, with 8 competitions associated to each of the two days. The list of competitions that ar part of the program are available here:
https://neurips.cc/Conferences/2020/CompetitionTrack
Machine learning competitions have grown in popularity and impact over the last decade, emerging as an effective means to advance the state of the art by posing well-structured, relevant, and challenging problems to the community at large. Motivated by a reward or merely the satisfaction of seeing their machine learning algorithm reach the top of a leaderboard, practitioners innovate, improve, and tune their approach before evaluating on a held-out dataset or environment. The competition track of NeurIPS has matured in 2020, its fourth year, with a considerable increase in both the number of challenges and the diversity of domains and topics. A total of 16 competitions are featured this year as part of the track, with 8 competitions associated to each of the two days. The list of competitions that ar part of the program are available here:
https://neurips.cc/Conferences/2020/CompetitionTrack
To ask questions please use rocketchat, available only upon registration and login.
Schedule
Fri, Dec 11th, 2020 @ 16:00 – 16:15 GMT
Opening - Competition Track Session
Katja Hofmann, Hugo Jair Escalante
Fri, Dec 11th, 2020 @ 16:15 – 16:35 GMT
3D+texture garment reconstruction challenge design (data, metrics, tracks, etc)
Hugo Bertiche
Fri, Dec 11th, 2020 @ 16:35 – 17:00 GMT
3D+texture garment reconstruction challenge results
Meysam Madadi
Fri, Dec 11th, 2020 @ 17:00 – 17:03 GMT
Opening the L2RPN challenge @ NeurIPS2020
Antoine Marot
Fri, Dec 11th, 2020 @ 17:03 – 17:13 GMT
Winning the L2RPN challenge
Antoine Marot
Fri, Dec 11th, 2020 @ 17:13 – 17:23 GMT
A L2RPN Winning approach
Jixiang LU
Fri, Dec 11th, 2020 @ 17:23 – 17:33 GMT
The Best L2RPN wining approach
Bo Zhou
Fri, Dec 11th, 2020 @ 17:33 – 17:40 GMT
L2RPN Post Challenge open questions
Antoine Marot
Fri, Dec 11th, 2020 @ 17:40 – 17:45 GMT
Closing and ceremony award
Antoine Marot
Fri, Dec 11th, 2020 @ 18:00 – 18:03 GMT
Introducing the Hide-and-Seek privacy challenge
James Jordon
Fri, Dec 11th, 2020 @ 18:03 – 18:18 GMT
The importance of synthetic data
James Jordon
Fri, Dec 11th, 2020 @ 18:18 – 18:28 GMT
Synthetic data in the healthcare setting
James Jordon
Fri, Dec 11th, 2020 @ 18:28 – 18:38 GMT
What we learned from the Hide-and-Seek privacy challenge
James Jordon
Fri, Dec 11th, 2020 @ 18:38 – 18:43 GMT
Closing remarks
James Jordon
Fri, Dec 11th, 2020 @ 19:00 – 19:16 GMT
Background on black box optimization (BBO)
Ryan Turner
Fri, Dec 11th, 2020 @ 19:16 – 19:27 GMT
BBO challenge platform with Valohai
Juha Kiili
Fri, Dec 11th, 2020 @ 19:27 – 19:32 GMT
Spotlight for 1st place (BBO challenge)
Ryan Turner, Alexander Cowen-Rivers
Fri, Dec 11th, 2020 @ 19:32 – 19:37 GMT
Spotlight for 2nd place (BBO challenge)
Ryan Turner, Jiwei Liu
Fri, Dec 11th, 2020 @ 19:37 – 19:42 GMT
Spotlight for 3rd place (BBO challenge)
Ryan Turner, Mikita Sazanovich
Fri, Dec 11th, 2020 @ 22:00 – 22:07 GMT
Opening the SpaceNet 7 Challenge @ NeurIPS2020
Adam Van Etten
Fri, Dec 11th, 2020 @ 22:07 – 22:15 GMT
Introduction to SpaceNet
Jake Shermeyer
Fri, Dec 11th, 2020 @ 22:15 – 22:25 GMT
The SpaceNet 7 Dataset
Jesus Martinez-Manso
Fri, Dec 11th, 2020 @ 22:25 – 22:30 GMT
The SpaceNet 7 Metric
Adam Van Etten
Fri, Dec 11th, 2020 @ 22:30 – 22:40 GMT
The Winners of SpaceNet 7
Adam Van Etten
Fri, Dec 11th, 2020 @ 22:40 – 22:45 GMT
SpaceNet 7 Closing and Future Plans
Adam Van Etten
Fri, Dec 11th, 2020 @ 23:00 – 23:05 GMT
introduction to the 2020 NeurIPS education challenge
Angus Lamb
Fri, Dec 11th, 2020 @ 23:05 – 23:20 GMT
Competition overview: motivation, impact, dataset, tasks
Angus Lamb
Fri, Dec 11th, 2020 @ 23:20 – 23:30 GMT
Competition results and insights
Jack Wang
Fri, Dec 11th, 2020 @ 23:30 – 23:35 GMT
Beyond the competition: what's next?
Jack Wang
Fri, Dec 11th, 2020 @ 23:35 – 23:45 GMT
Q&A and discussion
Jack Wang, Angus Lamb
Sat, Dec 12th, 2020 @ 00:00 – 00:05 GMT
Traffic Map Movies - An Introduction to the Traffic4cast Challenge
Sepp Hochreiter
Sat, Dec 12th, 2020 @ 00:05 – 00:10 GMT
The Traffic4cast Competition Design and Data
Michael Kopp
Sat, Dec 12th, 2020 @ 00:10 – 00:12 GMT
The Best Traffic4Cast Submissions
David Kreil, D Kreil
Sat, Dec 12th, 2020 @ 00:12 – 00:17 GMT
1st prize: Utilizing UNet for the Future Traffic Map Prediction - Traffic4cast highlight talk
Sungbin Choi
Sat, Dec 12th, 2020 @ 00:17 – 00:22 GMT
2nd prize: TLab: Traffic Map Movie Forecasting Based on HR-NET - Traffic4cast highlight talk
Fanyou Wu
Sat, Dec 12th, 2020 @ 00:22 – 00:27 GMT
3rd prize: Towards Good Practices of U-Net for Traffic Forecasting - Traffic4cast highlight talk
Jingwei Xu
Sat, Dec 12th, 2020 @ 00:27 – 00:32 GMT
Graph Ensemble Net and the Importance of Feature & Loss Function Design for Traffic Prediction - Traffic4cast highlight talk
Qi Qi
Sat, Dec 12th, 2020 @ 00:32 – 00:37 GMT
Uncertainty Intervals for Graph-based Spatio-Temporal Traffic Prediction - Traffic4cast highlight talk
tijs Maas
Sat, Dec 12th, 2020 @ 00:37 – 00:45 GMT
Traffic4cast Award Ceremony, Outlook, and Follow Up Challenges
D Kreil
Sat, Dec 12th, 2020 @ 01:00 – 01:15 GMT
The Hateful Memes Challenge: Competition Overview
Douwe Kiela
Sat, Dec 12th, 2020 @ 01:15 – 01:45 GMT
The Hateful Memes Challenge: Live award ceremony and winner presentations
Douwe Kiela
Competition Track Saturday
Hugo Jair Escalante, Katja Hofmann
Sat, Dec 12th @ 16:00 GMT – Sun, Dec 13th @ 01:45 GMT
Abstract: Second session for the competition program at NeurIPS2020.
Machine learning competitions have grown in popularity and impact over the last decade, emerging as an effective means to advance the state of the art by posing well-structured, relevant, and challenging problems to the community at large. Motivated by a reward or merely the satisfaction of seeing their machine learning algorithm reach the top of a leaderboard, practitioners innovate, improve, and tune their approach before evaluating on a held-out dataset or environment. The competition track of NeurIPS has matured in 2020, its fourth year, with a considerable increase in both the number of challenges and the diversity of domains and topics. A total of 16 competitions are featured this year as part of the track, with 8 competitions associated to each of the two days. The list of competitions that ar part of the program are available here:
https://neurips.cc/Conferences/2020/CompetitionTrack
Machine learning competitions have grown in popularity and impact over the last decade, emerging as an effective means to advance the state of the art by posing well-structured, relevant, and challenging problems to the community at large. Motivated by a reward or merely the satisfaction of seeing their machine learning algorithm reach the top of a leaderboard, practitioners innovate, improve, and tune their approach before evaluating on a held-out dataset or environment. The competition track of NeurIPS has matured in 2020, its fourth year, with a considerable increase in both the number of challenges and the diversity of domains and topics. A total of 16 competitions are featured this year as part of the track, with 8 competitions associated to each of the two days. The list of competitions that ar part of the program are available here:
https://neurips.cc/Conferences/2020/CompetitionTrack
To ask questions please use rocketchat, available only upon registration and login.
Schedule
Sat, Dec 12th, 2020 @ 16:00 – 16:05 GMT
Introducing EfficientQA: Open domain question answering with memory constraints as a testbed for language understanding and knowledge representations
Tom Kwiatkowski
Sat, Dec 12th, 2020 @ 16:05 – 16:25 GMT
Track winner presentations:
Tom Kwiatkowski
- smallest question answering system with 25% accuracy - best performing system smaller than 500 Mb - best performing system smaller than 6Gb - best performing system overall
Sat, Dec 12th, 2020 @ 16:25 – 16:45 GMT
Showdown against trivia experts
Jordan Boyd-Graber
Five top human teams of trivia experts took on the competition’s baseline systems for the opportunity to take on the computer systems in each of the competition’s divisions. The team of humans will compete against the computer on thirty questions from the test set. We will present highlights from the preliminary competition as well as the final showdown between computer systems and the human teams.
Sat, Dec 12th, 2020 @ 17:00 – 17:06 GMT
Predicting Generalization in Deep Learning (PGDL): Opening remark
Yiding Jiang
Sat, Dec 12th, 2020 @ 17:06 – 17:16 GMT
Keynote speech: Sanjeev Arora (PGDL)
Sanjeev Arora, Yiding Jiang
Sat, Dec 12th, 2020 @ 17:16 – 17:17 GMT
Introduction to winning team (PGDL)
Yiding Jiang
Sat, Dec 12th, 2020 @ 17:17 – 17:27 GMT
Winning team presentation: On Representations and Generalization (PGDL)
Yiding Jiang, Parth Natekar, Manik Sharma
Sat, Dec 12th, 2020 @ 17:27 – 17:28 GMT
Introduction to Runner up 1 (PGDL)
Yiding Jiang
Sat, Dec 12th, 2020 @ 17:28 – 17:33 GMT
Runner up presentation: Robustness to Augmentations as a Generalization Metric (PGDL)
Sumukh Aithal K
Sat, Dec 12th, 2020 @ 17:33 – 17:34 GMT
Introduction to Runner up 2 (PGDL)
Yiding Jiang
Sat, Dec 12th, 2020 @ 17:34 – 17:39 GMT
Runner up presentation: Ranking generalization via smoothness of latent graphs (PGDL)
Carlos Lassance
Sat, Dec 12th, 2020 @ 17:39 – 17:44 GMT
Closing remark (PGDL)
Yiding Jiang
Sat, Dec 12th, 2020 @ 18:00 – 18:15 GMT
Openning and design of the INTERPRET challenge @ NeurIPS2020
Wei Zhan
Sat, Dec 12th, 2020 @ 18:15 – 18:35 GMT
Winner talks of INTERPRET challenge
Hengbo Ma
Sat, Dec 12th, 2020 @ 18:35 – 18:45 GMT
Analysis, research opportunities and closing of INTERPRET challenge
Liting Sun
Sat, Dec 12th, 2020 @ 19:00 – 19:05 GMT
NLC2CMD Competition Organizers: Introduction, Problem Description, CLAI
Mayank Agarwal
Sat, Dec 12th, 2020 @ 19:05 – 19:15 GMT
NLC2CMD Competition Keynote: Tellina
Victoria Lin, Mayank Agarwal, Tathagata Chakraborti
Sat, Dec 12th, 2020 @ 19:15 – 19:20 GMT
NLC2CMD Competition Organizers: Metrics, Data, Tracks
Mayank Agarwal
Sat, Dec 12th, 2020 @ 19:20 – 19:22 GMT
NLC2CMD Competition Participant Team: AINixClaiSimple
David Gros
Sat, Dec 12th, 2020 @ 19:22 – 19:24 GMT
NLC2CMD Competition Participant Team: coinse-team
Juyeon Yoon
Sat, Dec 12th, 2020 @ 19:24 – 19:26 GMT
NLC2CMD Competition Participant Team: AICore
Kangwook Lee
Sat, Dec 12th, 2020 @ 19:26 – 19:28 GMT
NLC2CMD Competition Participant Team: magnum
Quchen Fu
Sat, Dec 12th, 2020 @ 19:28 – 19:30 GMT
NLC2CMD Competition Participant Team: Hubris
Jaron Maene
Sat, Dec 12th, 2020 @ 19:30 – 19:32 GMT
NLC2CMD Competition Participant Team: jb
Denis Litvinov
Sat, Dec 12th, 2020 @ 19:32 – 19:45 GMT
NLC2CMD Competition Organizers: Results (Live)
Kartik Talamadupula
Sat, Dec 12th, 2020 @ 22:00 – 22:03 GMT
Introduction to AIDO
Liam Paull
Sat, Dec 12th, 2020 @ 22:03 – 22:05 GMT
Short Scientific Talk (AIDO)
Luigi Di Lillo
Sat, Dec 12th, 2020 @ 22:05 – 22:13 GMT
Advanced Perception League
Liam Paull
Sat, Dec 12th, 2020 @ 22:13 – 22:19 GMT
Intro to Urban League (includes highlights from semifinals)
Liam Paull
Sat, Dec 12th, 2020 @ 22:19 – 22:34 GMT
Live robot competition (LF, LFP, lFVM)
Liam Paull
Sat, Dec 12th, 2020 @ 22:34 – 22:40 GMT
Interviews with winners
Liam Paull
Sat, Dec 12th, 2020 @ 22:40 – 22:43 GMT
Conclusions and Wrap up
Liam Paull
Sat, Dec 12th, 2020 @ 23:00 – 23:05 GMT
Introduction - Flatland
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:05 – 23:15 GMT
Flatland Competition Design & Results
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:15 – 23:19 GMT
Winner Talks : Team An Old Driver
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:19 – 23:23 GMT
Winner Talks : Team JBR_HSE
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:23 – 23:27 GMT
Winner
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:27 – 23:42 GMT
"Real world applications of Flatland" : Panel Discussion with SBB, DeutschBahn, SNCF
Sharada Mohanty
Sat, Dec 12th, 2020 @ 23:42 – 23:45 GMT
Concluding Remarks
Sharada Mohanty
Sun, Dec 13th, 2020 @ 00:00 – 00:03 GMT
Introduction - Procgen
Sharada Mohanty
Sun, Dec 13th, 2020 @ 00:03 – 00:12 GMT
Introduction to the Procgen Benchmark
Karl Cobbe
Sun, Dec 13th, 2020 @ 00:12 – 00:20 GMT
NeurIPS 2020 Procgen Challenge Design
Sharada Mohanty
Sun, Dec 13th, 2020 @ 00:20 – 00:25 GMT
Winner Announcements & Analysis of top submissions
Sharada Mohanty
Sun, Dec 13th, 2020 @ 00:25 – 00:44 GMT
Sample Efficiency & Generalization in RL : An assortment of tricks (talks by top participants)
Sharada Mohanty
Sun, Dec 13th, 2020 @ 00:44 – 00:45 GMT
Concluding Remarks
Sharada Mohanty
Sun, Dec 13th, 2020 @ 01:00 – 01:45 GMT
Introduction and results of the 2020 MineRL Competition
William Guss, Stephanie Milani, Nicholay Topin