Submission

Call for submissions

We invitie submissions in topics related to (but not at all limited to) the following topics:

  • Geometric video learning

  • Graph-based video representations

  • Disentangled video representations

  • Self-supervised learning in videos

  • Spatio-temporal human and object relation representation

  • Video learning with prior or symbolic knowledge

We accept two types of submissions:

Unpublished work: Papers that have not been published or accepted for publication in a similar form. This also includes papers currently under review at other venues. Please use the 8-page single blind format of ICCV for this kind of submission. Format instructions are available here. Note: no anonymization is needed.

Published work: Papers that been been published or accepted for publication in a recent venue, including conferences and journals. This also includes papers at the main conference of ICCV 2021.

Accepted papers will not appear in the conference proceedings. For unpublished works, we encourage authors to post their papers on arXiv, so that we can share the link on the workshop website.

All papers will be presented during oral sessions covering the entire day.

Ethical considerations

While the workshop focuses on understanding video content from a computer vision and machine learning content, the goal on the horizon is to understand what is going on when, where, and perhaps even why. Such an in-depth visual understanding has a wide range of potential applications. Video understanding can help towards improving elderly care (e.g. by identifying whether elderly people have fallen) and avoid poaching (e.g. using cameras on drones). On the other hand, it can also be used for surveillance and automated military drone applications. To foster open discussion, we ask for a Broader Impact statement (e.g. in the conclusions of the paper or in the talk), following the example set by NeurIPS 2020.

Important dates

Submission deadline: August 27th

Notification to authors: September 24th

Submission instructions

All papers have to be submitted through the CMT website of the workshop.

Full link: https://cmt3.research.microsoft.com/SRVU2021/Submission/Index