Robust Vision Challenge

The Robust Vision Challenge 2022 was a virtual-only full day event held on
October 23rd 2022 in conjunction with ECCV 2022 in Tel Aviv.

Schedule for 2022-10-23

Morning Session: 09h-11h30 UTC+3

  • 09h00-09h15: Introduction / Announcement of RVC Winners
  • 09h15-10h00: Invited Talk 1: Olga Russakovsky
  • 10h00-10h20: 1st Place (Obj. Det.)
  • 10h25-10h45: 2nd Place (Obj. Det.)
  • 10h45-11h05: 1st Place (Semantic)
  • 11h10-11h30: 2nd Place (Semantic)

Afternoon Session: 12h30-17h00 UTC+3

  • 12h30-13h15: Invited Talk 2: Judy Hoffman
  • 13h15-13h35: 1st Place (Stereo)
  • 13h40-14h00: 2nd Place (Stereo)
  • 14h00-14h20: 1st Place (Flow)
  • 14h25-14h45: 2nd Place (Flow)
  • 14h50-15h10: 1st Place (Depth)
  • 15h10-15h30: 2nd Place (Depth)
  • 15h30-16h15: Invited Talk 3: Chelsea Finn
  • 16h15-17h00: Open Discussion and Closing Remarks

Robust Vision Challenge 2022

The increasing availability of large annotated datasets such as Middlebury, PASCAL VOC, ImageNet, MS COCO, KITTI and Cityscapes has lead to tremendous progress in computer vision and machine learning over the last decade. Public leaderboards make it easy to track the state-of-the-art in the field by comparing the results of dozens of methods side-by-side. While steady progress is made on each individual dataset, many of them are limited to specific domains. KITTI, for example, focuses on real-world urban driving scenarios, while Middlebury considers indoor scenes. Consequently, methods that are state-of-the-art on one dataset often perform worse on a different one or require substantial adaptation of the model parameters.

The goal of this challenge is to foster the development of vision systems that are robust and consequently perform well on a variety of datasets with different characteristics. Towards this goal, we propose the Robust Vision Challenge, where performance on several tasks (eg, reconstruction, optical flow, semantic/instance segmentation, single image depth prediction) is measured across a number of challenging benchmarks with different characteristics, e.g., indoors vs. outdoors, real vs. synthetic, sunny vs. bad weather, different sensors. We encourage submissions of novel algorithms, techniques which are currently in review and methods that have already been published.

Questions? Please see the Submit Results page for anwsers regarding the RVC rules.


Prizes for winner and the runner-up of each of the seven challenges:

1st Place: TBD

2nd Place: TBD

Presentation at our
ECCV 2022 Workshop

Invitation to publish in our
IJCV Special Issue

Challenges

RVC 2022 features five challenges: stereo, optical flow, single image depth prediction, object detection, and semantic segmentation. Participants are free to submit to a single challenge or to multiple challenges. For each challenge, the results of a single model has to be submitted to all benchmarks (indicated with an x below).

Stereo
Flow
Depth
Obj. Det.
Semantic

Important Dates

Juni 15, 2022 Training data and development kit released:
https://github.com/ozendelait/rvc_devkit
July 15, 2022 RVC Submission system is online
October 3, 2022 Submission deadline (23h59 CEST/UTC+2 == 2:59pm PDT)
October 14, 2022 Report deadline (23h59 UTC+2)
October 23, 2022 Robust Vision Challenge 2022 Workshop at ECCV 2022 (Virtual Fullday)

Sponsors

Gold Sponsors



eXTReMe Tracker