[robotics-worldwide] [meetings] CVPR 2017 Workshop on Continuous and Open-Set Learning

Erik Rodner Erik.Rodner at uni-jena.de
Sun May 7 12:54:52 PDT 2017


We invite 1-page abstract submissions to our CVPR 2017 Workshop on
Continuous and Open-Set Learning.

Deadline: 31st of May, 2017

Website: https://urldefense.proofpoint.com/v2/url?u=https-3A__erodner.github.io_continuouslearningcvpr2017_&d=DwIFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=cve0_W_w-Iu2V5a77WiS2yjyvIoM0H-lRUrIB6kcupo&s=9scstr12PQbZ05OlsnmO2l83VEYex8RpuPZMDGPgiUg&e= 

Workshop Date: 26th of July, 2017

Recent breakthroughs in our community have relied on the availability of
large representative datasets for training. However, the implicit
assumption imposed in the majority of our today’s techniques is a static
closed world. These assumptions rarely hold in many application areas,
instead, the set of semantic concepts and relevant tasks is dynamically
changing - even on a daily basis. The assumption of a closed and static
world is therefore one of the major obstacles when building intelligent
systems that learn continuously, adaptively, and actively.

The workshop tries to bridge one of the gaps between computer vision
research and AI goals by focusing on different aspects of continuous and
open-set learning.


Erik Rodner (Corporate Research and Technology, Carl Zeiss AG)

Alexander Freytag (Corporate Research and Technology, Carl Zeiss AG)

Christoph Lampert (IST Austria)

Terrance E. Boult ((University of Colorado, Colorado Springs, USA)

Joachim Denzler (University of Jena)

The following topics will be central to the workshop:


   Dealing with partially unknown, open, or dynamically increasing label
   spaces (probabilistic models, possibility for rejection, novelty detection,

   Continuous, online, and incremental learning (at level of instances,
   classes, common-sense knowledge, and representations)

   Active acquisition and annotation of new data with humans in the loop
   (curriculum learning, active learning, etc.)

   Transfer learning and domain adaptation in continuous and open-set
   learning scenarios

   Active data discovery in explorative data science and large-scale
   microscopy data

   Benchmarking success in continuous and open-set learning scenarios

Abstract submissions are limited to one page and need to be in CVPR17 format
<https://urldefense.proofpoint.com/v2/url?u=http-3A__cvpr2017.thecvf.com_submission_main-5Fconference_author-5Fguidelines&d=DwIFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=cve0_W_w-Iu2V5a77WiS2yjyvIoM0H-lRUrIB6kcupo&s=lqKwwwjPda4SlaRicoEREioIbXmgZ8dW9P5wRrvP__w&e= >.
The abstract will not appear in any proceedings and if accepted only appear
online on this page (if authors like). Our workshop is not meant as a
publication venue, but rather a real meeting, where you learn about people
interested in the same area and find the next cooperation partners for your
future project.

Accepted abstracts will be presented in a quick 5min talk and a poster. We
also welcome submissions of industrial partners interested in the topic and
willing to present their application area. Furthermore, if you want to
present your next proposal idea and you are looking for cooperation
partners, you are also very much invited to submit an abstract.

We are looking forward to your submissions, see you at CVPR 2017.

Erik Rodner (on behalf of the other organizers)

More information about the robotics-worldwide mailing list