Welcome

Contact

Philipp Krähenbühl
353 Serra Mall
Room 286
Stanford CA 94305-9025
email: philkr(at)stanford.edu
curriculum vitae

Research

I am a Ph.D. candidate in the Department of Computer Science at Stanford University. I am working with Vladlen Koltun and the Stanford Virtual Worlds Group. My research interests are computer graphics in particular character animation.

Previous research:

• Video Retargeting at ETH Zurich, with Prof. Markus Gross
• Fold.it at University of Washington, with Prof. Zoran Popovic and Adrien Treuille
• “SWPS3-fast multi-threaded vectorized Smith-Waterman for IBM Cell/B.E.“ at ETH Zurich, with Prof. Gaston Gonnet and Christophe Dessimoz
• EiffelMedia at ETH Zurich, with Prof. Bertrand Meyer and Till Bay

Teaching

• CSE 524 - Parallel Algorithms with Prof. Larry Snyder
• Introduction to Programming with Prof. Bertrand Meyer

Publications

	Philipp Krähenbühl, Manuel Lang, Alexander Hornung and Markus Gross Retargeting of Streaming Video Proceedings of ACM SIGGRAPH Asia (Yokohama, Japan, December 16-19, 2009), ACM Transactions on Graphics [PDF] [Video] We present a novel, integrated system for content-aware video retargeting. A simple and interactive framework combines key frame based constraint editing with numerous automatic algorithms for video analysis. This combination gives content producers high level control of the retargeting process. The central component of our framework is a non-uniform, pixel-accurate warp to the target resolution which considers automatic as well as interactively defined features. Automatic features comprise video saliency, edge preservation at the pixel resolution, and scene cut detection to enforce bilateral temporal coherence. Additional high level constraints can be added by the producer to guarantee a consistent scene composition across arbitrary output formats. For high quality video display we adopted a 2D version of EWA splatting eliminating aliasing artifacts known from previous work. Our method seamlessly integrates into postproduction and computes the reformatting in realtime. This allows us to retarget annotated video streams at a high quality to arbitary aspect ratios while retaining the intended cinematographic scene composition. For evaluation we conducted a user study which revealed a strong viewer preference for our method.
	Philipp Krähenbühl, Manuel Lang, Alexander Hornung and Markus Gross Art Directable Retargeting for Streaming Video Bachelor Thesis, ETH Zurich, February 2009 [PDF] We present a novel framework for content-aware and art-directable video retargeting. A simple and interactive workflow combines key frame based constraint editing with numerous automatic algorithms for video analysis. This combination gives content producers high level control of the retargeting process. The central component of our framework is a non-uniform, pixelaccurate warp to the target resolution which considers automatic as well as interactively defined features. Automatic features comprise video saliency, edge preservation at the pixel resolution, and a scene cut detection to enforce bilateral temporal coherence. Additional high level constraints concerning scene composition can be added by the producer. For high quality video display we adopted a 2D version of EWA splatting eliminating aliasing artifacts known from previous methods. Our method seamlessly integrates into postproduction and computes the reformatting in realtime. This allows us to retarget annotated video streams to arbitary devices while retaining the intended cinematographic scene composition.
	Adam Szalkowski, Christian Ledergerber, Philipp Krähenbühl and Christophe Dessimoz SWPS3 - fast multi-threaded vectorized Smith-Waterman for IBM Cell/B.E. and x86/SSE2 BMC Research Notes 2008, 1:107 [Project] We present swps3, a vectorized implementation of the Smith-Waterman local alignment algorithm optimized for both the Cell/BE and x86 architectures. The paper describes swps3 and compares its performances with several other implementations. Our benchmarking results show that swps3 is currently the fastest implementation of a vectorized Smith-Waterman on the Cell/BE, outperforming the only other known implementation by a factor of at least 4: on a Playstation 3, it achieves up to 8.0 billion cell-updates per second (GCUPS). Using the SSE2 instruction set, a quad-core Intel Pentium can reach 15.7 GCUPS. We also show that swps3 on this CPU is faster than a recent GPU implementation. Finally, we note that under some circumstances, alignments are computed at roughly the same speed as BLAST, a heuristic method. The Cell/BE can be a powerful platform to align biological sequences. Besides, the performance gap between exact and heuristic methods has almost disappeared, especially for long protein sequences.