Skip to content

GitLab

M
MLDL_FZJ_Wiki
Last edited by Mehdi Cherti
Page history
This is an old version of this page. You can view the most recent version or browse the history.

Juelich Problems

Based on existing experience in creating, hosting and advertising plant science data challenges, thus data science problems and according benchmark data, we aim on connecting working groups who are interested to follow this concept and in the long term to organize a joint workshop and publication on 'Juelich Problems'.

Workshop establishing challenge datasets 'Juelich Problems'

  • When: Thursday, 8 November 2018, 10:00-12:00h
  • Where: IBG-2 seminar room (building 06.2, room 406)

Scope

Over the last few years we had excellent experiences with publishing datasets to draw the interest of the computer vision and machine learning community to our applications (see datasets-impact). This not only led to increased visibility (>1600 downloads, >140 citations), but most importantly to highly increased accuracy of available solutions to our image analysis problems – because others worked on them ‘for free’.

With this workshop, we want to

  • Share our experiences, including dos and don’ts
  • Give advice and discuss how to promote our Juelich application problems successfully (offer image data together with ground truth, organize challenges and workshops, publish overview papers and problem statement papers as well as websites, follow F.A.I.R. principles etc.)
  • Find people (you?) interested in publishing their data in a similar way to tap its full potential
  • Join forces and discuss further steps

Slides and Minutes

Intro

Spotlight Talks