CVPR 2015 Workshop
June 11, 2015 (PM Only), Room 204
Boston, MA, USA


Broad access to image and video datasets has been responsible for much of the progress in computer vision recognition problems over the last decade. These common benchmarks have played a leading role in transforming recognition research from a black art into an experimental science. Progress, however, has stagnated; although datasets continue to grow, they are developed and annotated in isolation: e.g., a collection of sporting activities, a set of objects in images, etc. These isolated datasets suffer from task and domain-specific bias, and knowledge transfer across them is extremely limited. This workshop intends to reinvigorate the datasets conversation in the computer vision community in an open-forum of discussion and collaboration. 

Our primary goal is to, as a community, set a direction toward which we can collectively work in creating a federated dataset infrastructure that will allow for transparency across individual datasets and annotations, experimental benchmarks with community-set corpora and metrics, and a web-based infrastructure to cultivate continued developed of computer vision datasets. We will discuss dataset availability, dataset creation and annotation issues, cross-dataset challenge problems, as well as dataset infrastructure issues.

Sponsored by