Visual Crowds for Geo- and Environmental Sciences Conference Heidelberg

In this workshop, we want to tie two threads of research together: human computation is an important factor in both geoscience and computer vision research. In geosciences, OpenStreetMap is one of the major projects , but this is only one of numerous recent advances. In environmental sciences huge amounts of data are collected, such as high-speed recordings of fluids, thousands of meters of glacier ice drilling cores or satellite data depicting e.g. sand transport in the Sahara desert. In computer vision everything started out with the ESP Game, Peekaboom and the LabelMe tools. Today, crowdsourcing is used in almost every discipline dealing with image data. However, these communities rarely talk with each other. Publications are mainly focused at the disciplines’ top journals with little to no overlap.

Heidelberg University has been an international leader in the usage, analysis and improvement of OpenStreetMap in many domains and applications, reflected e.g. by its graduate school CrowdAnalyser dedicated to this topic. The Heidelberg Center for the Environment (HCE) connects a wide array of research fields ranging from natural sciences to social and cultural sciences. The Heidelberg Collaboratory for Image Processing is one of the largest institutes for computer vision in Germany with more than 80 researchers working in all subfields ranging from machine learning over early vision to scene parsing.

Date: Tuesday, December 9, 2014 @ Heidelberg University

More information on the Workshop and participation can be found here


Daniel Kondermann

Jamal Jokar

Bernd Jähne

Alexander Zipf

Werner Aeschbach-Hertig