Thursday, September 8, 2016

13:30 Welcome and introduction
  Jochem Herrmann, Adimec, Eindhoven (The Netherlands), President EMVA
Bernd Jähne, HCI, Heidelberg University (Germany), EMVA Board member
13:45 Feature extraction
  Visual tracking - the transition from hand-crafted features to deep features
Prof. Dr. Michael Felsberg, Linköping University, Sweden
  Fast unsupervised detection of segments and curves in images and applications
Rafael Grompone and Prof. Jean-Michel Morel, CMLA, Université Paris-Saclay and CNS Cachan, France
  Time-causal and time-recursive receptive fields for invariance and covariance under natural image transformations
Prof. Dr. Tony Lindeberg, KTH Royal Institute of Technology, Stockholm, Sweden
15:15 Coffee break / demos / poster
16:00 Segmentation and 3-D reconstruction
  Spline-based models for image segmentation
Virginie Uhlmann, EPFL Lausanne, Switzerland
  Multicuts – an optimal way to turn superpixels into regions
Dr. Ullrich Köthe, HCI, Heidelberg University, Germany
  Direct methods for camera localisation and 3-D reconstruction
Christian Kerl and Prof. Dr. Daniel Cremers, Computer Vision Group, Techn. University Munich, Germany
  Reconstruction via detection: Efficient and automatic reconstruction from unorganized 3D scans
Tolga Birdal, Techn. University Munich, and Siemens, Munich, Germany
18:00 Poster and demo teaser
18:30 Evening get-together with finger food, software demonstrations, poster presentations and discussions in small groups(open end)

Friday, September 9, 2016

09:00 Image compression with diffusion-based inpainting
Prof. Dr. Joachim Weickert, Saarland University, Saarbrücken, Germany
  Subdivision surfaces in computer vision
Dr. Andrew Fitzgibbon, Microsoft Research, Cambridge, United Kingdom
  Performance evaluation in computer vision
Katrin Honauer, HCI, Heidelberg University, Germany
10:30 Coffee break / demos / poster
11:15 Learning
  Recent advances in deep learning for object detection and image segmentation
Prof. Frédéric Jurie, Université de Caen, France
  End2end learning of CNNs and CRFs with application to stereo
Prof. Dr. Thomas Pock, Techn. University Graz, Austria
  Compositional visual learning and retrieval
Prof. Dr. Björn Ommer, HCI, Heidelberg University, Germany
12:45 Closing remarks
13:00 Farewell lunch, continuing software demonstrations, poster presentations and discussions in small groups
End at approx. 14:00 h