Alumni Updates, Mike Honeck and Federico Perazzi

Mike Honeck, class of 2011, was part of a team that just presented some nice work at the Society for Information Display Conference in San Jose, CA last month.

A Tracked Automultiscope 3D Tabletop

Quinn Smithwick (DRLA), Michael Honeck (WDI)

Society for Information Display Week 2015 Conference

May 31, 2015 – San Jose, CA

To make a large wide field-of-view 3D tabletop display, which does not require viewers donning 3D glasses, we investigate an autostereoscopic extension of anamorphic projection. We created a 42” autostereoscopic tabletop using a commercial lenticular display with viewer tracking and multiscopic viewpoint re-projection to overcome the lenticular display’s inherent limitations, such as horizontal-parallax-only perspectives, limited fieldof-view and repeated view zones. A single viewer can observe full-parallax synthetic 3D objects on the large autostereo tabletop over a wide field of view (120°) without the need for 3D glasses.

 

---

 

Federico Perazzi, class of 2010, and his team presented on Efficient Salient Foreground Detection for Images and Video using Fiedler Vectors at the 4th Workshop on Intelligent Camera Control, Cinematography and Editing in  Zürich, Switzerland in early May 2015.

Efficient Salient Foreground Detection for Images and Video using Fiedler Vectors

Federico Perazzi (DRZ/ETH Joint PhD), Olga Sorkine-Hornung (ETH Zurich), Alexander Sorkine-Hornung (DRZ)

4th Workshop on Intelligent Camera Control, Cinematography and Editing

May 4, 2015 – Zürich, Switzerland

Automatic detection of salient image regions is a useful tool with applications in intelligent camera control, virtual cinematography, video summarization and editing, evaluation of viewer preferences, and many others. This paper presents an effective method for detecting potentially salient foreground regions. Salient regions are identified by eigenvalue analysis of a graph Laplacian that is defined over the color similarity of image superpixels, under the assumption that the majority of pixels on the image boundary show nonsalient background. In contrast to previous methods based on graph-cuts or graph partitioning, our method provides continuously valued saliency estimates with complementary properties to recently proposed color contrastbased approaches. Moreover, exploiting discriminative properties of the Fiedler vector, we devise an SVM-based classifier that allows us to determine whether an image contains any salient objects at all, a problem that has been largely neglected in previous works. We also describe how the per-frame saliency detection can be extended to improve spatiotemporal coherence on video sequences. Extensive evaluation on several datasets demonstrates and validates the state-of-the-art performance of the proposed method.

 

Impressive work, Mike and Federico!

 

If any ETC alumni would like to give any updates on what they are working on, please contact MaryCatherine with the information.

---