DThree19

Is your deep detector safe? verification and certification of neural networks

CVPR Workshop, 17 June 2019, Long Beach, CA, US

Home Submissions Organisers Speakers Challenge Program Contact

Submissions


    This is a half-day workshop with the primary objective of bringing this crucial aspect of safety of deep networks into the center of computer vision discussion. Keeping this in mind, we solicit invited contributions from the speakers as well as posters from the participants; we will select 5 posters for 3+1 minutes of spot-light presentation. There will also be an industrial demo session, where participants from industry would be demonstrating their current deep-learning based computer vision. We expect the participants to carry their own hardware (e.g., DrivePX2 or performed on a laptop). At least 4 display screens would be provided to run these demos in a loop.

Guidelines for submissions


    We solicit short papers and posters for the workshop. Submissions will be considered in following categories:

  • Abstract papers - 2 pages (without references)
  • Poster papers - 6 pages (with references)

    • All submissions should be formatted according to the LNCS submission guidelines, and should be submitted as Portable Document Format (PDF) files using the Easychair submission site DThree19. Submissions will be evaluated on quality as well as relevance to the workshop's theme. The deadline for initial submission is 10 March, 2019 at 23:59 PDT.

    All accepted papers will be presented during a poster session, and a subset will be selected for a short oral presentation. We also plan to have post-proceedings publication for all accepted papers; the details would be updated by late March, 2019.

    To foster dialogue and collaboration between researchers, this workshop will be designed to be as interactive as possible, with organized discussions and informal breakout sessions. As such, a large part of this effort is to bring together a community of robotic researchers who are working in the area of formal methods and computer vision in the context of modern machine learning. Ample time will be provided for networking and informal discussions.

Important Dates

  • Submission Deadline: 10 March, 2019
  • Acceptance Notification: 25 March, 2019
  • Camera-ready Submission: 31 March, 2019
  • Workshop date: 17 June, 2019