Rogerio Schmidt Feris

Research Staff Member

PhD, University of California, Santa Barbara, 2006

rogerioferis-AT-yahoo.com (personal)
rsferis-AT-us.ibm.com (work)

Short Biography:

I am currently a research scientist at IBM T.J. Watson Research Center, New York. I received a PhD in computer science from the University of California, Santa Barbara, an MS in computer science from University of Sao Paulo, Brazil, and a BS in computer engineering from the Federal University of Rio Grande, Brazil. My research interests include computer vision, graphics, computational photography, and machine learning. My publications have appeared in major computer vision/graphics conferences and journals, including ICCV, CVPR, SIGGRAPH, and PAMI. Throughout my graduate studies, I did research internships at Microsoft Research, Mitsubishi Electric Research Labs (MERL), and IBM Research. In 2002, my MS thesis on wavelet subspace tracking was awarded in a nationwide contest in Brazil. In 2005, I was named by IBM as “Emerging Leader in Multimedia”, an award given to eight students selected from top universities in US. I am a member of the ACM and IEEE and have served as a program committee member of major computer vision conferences such as CVPR, ICCV, and ECCV.

As an Adjunct Professor at Columbia University, I co-taught a course on Automatic Video Surveillance (Spring 2008). I am also an affiliate assistant professor at University of Washington (EE department), serving as a PhD co-advisor since July 2008. In addition to working on core research, I had a one-year assignment at IBM Global Technology Services as a senior software engineer to help the productization of the IBM Smart Surveillance System.

PhD Thesis:

Detection and Modeling of Depth Discontinuities with Lighting and Viewpoint Variation
Thesis committee: Matthew Turk (advisor), Ramesh Raskar, Steve Seitz, Yuan-Fang Wang, and Tobias Hollerer

Research Interests: My interests span a wide range of topics in computer vision and graphics, with emphasis on visual surveillance, intelligent user interfaces, and digital photography applications.
Specific projects include automated video analysis for surveillance, flash photography for image analysis and rendering, object classification, and face/gesture recognition

Research Internships :

IBM Research - I worked with Ying-Li Tian (summer 2005) in the Exploratory Computer Vision Group , Hawthorne, New York (IBM T. J. Watson Research Center). Project: Multiview Face Detection for EC-ASSIST, a Darpa project involving IBM, Georgia Tech, MIT and UC-Irvine.

Mitsubishi Electric Research Labs - I worked with Dr. Ramesh Raskar at MERL in Cambridge, Massachussets. Project: A non-photorealistic ('NPR') camera, a self-contained device that makes use of a multi-flash technique to render cartoon-like, easy to understand images of complex scenes.

Microsoft Research - I worked in the Media Presence group in San Francisco (Microsoft Bay Area Research Center) with Dr. Jim Gemmell and Dr. Kentaro Toyama. Project: GazeMaster, a videoconferencing system that uses computer vision and graphics methods to provide gaze awareness and sense of space. Look at my final evaluation here.

Research (Click here for a complete set of publications)

	Characterizing the Shadow Space of Camera-Light Pairs (with Daniel Vaquero, Ramesh Raskar, and Matthew Turk) IEEE Conference on Computer Vision and Pattern Recognition (CVPR'08). Click here for more recent work related to multi-flash imaging.
	IBM Smart Surveillance Solution (with the Exploratory Computer Vision Group at IBM) I'm currently leading face analytics and object classification research for this project. See a VIDEO DEMO of our system and our face capture method.
	Discontinuity Preserving Stereo with Small Baseline Multi-Flash Illumination (with Longbin Chen, Matthew Turk, Ramesh Raskar, and Karhan Tan ) Accepted for publication at International Conference on Computer Vision (ICCV'05), oral presentation (3.7% acceptance)
	Automatic Human Facial Illustrations with Variable Illumination (with Alex Olwal and Matthew Turk) This project aims to combine computer vision and graphics for interactive and automatic human facial illustrations. We will demonstrate it in the Interactive FogScreen demo, SIGGRAPH Emerging Technologies, 2005.
	A Non-Photorealistic Camera: Depth Edge Detection and Stylized Rendering with Multi-Flash Imaging (with Ramesh Raskar, Karhan Tan, Jingyi Yu and Matthew Turk) Accepted for publication in *ACM SIGGRAPH 2004* (see a VIDEO DEMO of our work) Also accepted for Siggraph Emerging Technologies, 2004
	Exploiting Depth Discontinuities for Vision-Based Fingerspelling Recognition (with Matthew Turk, Ramesh Raskar, Karhan Tan and Gosuke Ohashi) IEEE Workshop on Real-time Vision for Human-Computer Interaction (in conjunction with CVPR'04), Washington DC, USA, 2004. Extended version to appear as a book chapter, Springer-Verlag press.
	Specular Reflection Reduction with Multi-Flash Imaging (with Ramesh Raskar, Karhan Tan and Matthew Turk) IEEE Brazilian Symposium on Computer Graphics and Image Processing (Sibgrapi'04), Curitiba, Brazil, 2004. Also accepted as a Siggraph poster, 2004. Extended version to appear in Journal of Brazilian Computer Society.
	Shape Enhanced Surgical Visualizations with Multi-Flash Imaging (with Karhan Tan, James Kobler, Paul Dietz and Ramesh Raskar International Conference on Medical Imaging Computing and Computer Assisted Intervention (Miccai'2004), France, 2004
	Manifold Based Analysis of Facial Expressions (with Changbo Hu, Ya Chang and Matthew Turk) IEEE Workshop on Face Processing in Video (in conjunction with CVPR'04), Washington DC, USA, 2004. See our VIDEO demonstration. Extended version in Image and Vision Computing Journal.
	Real-time View-based Face Alignment Using Active Wavelet Networks (with Changbo Hu and Matthew Turk) IEEE Workshop on Analysis and Modeling of Faces and Gestures (in conjunction with ICCV'03), Nice, France, 2003
	Active Wavelet Networks for Face Alignment (with Changbo Hu and Matthew Turk) British Machine Vision Conference (BMVC'03), Norwich, UK, 2003
	GazeMaster: Gaze-aware Videoconferencing (with Jim Gemmell and Kentaro Toyama) This is the project I worked at Microsoft Research (follow the link for publications)
	Head Pose Tracking using Incremental Focus of Attention (with Jim Gemmell and Kentaro Toyama) I extended Kentaro Toyama's tracker to achieve subpixel accuray (see VIDEO)
	Real-time Detection, Alignment and Recognition of Human Faces (with Changbo Hu and Matthew Turk) We developed a fully automatic recognition system, with evaluation on Feret database.
	Hierarchical Wavelet Networks for Facial Feature Localization (with Jim Gemmell, Kentaro Toyama and Volker Krueger) International Conference on Automatic Face and Gesture Recognition (FG'02), Washington, DC, 2002
	Efficient Real-Time Face Tracking in Wavelet Subspace (with Volker Krueger and Roberto Cesar) IEEE Workshop on Recognition, Analysis, and Tracking of Faces and Gestures in Real-time Systems (in conjunction with ICCV'01), Vancouver, Canada, 2001. Also accepted for Real-Time Imaging Journal This is part of my MS thesis work (which was awarded in a Brazilian nationwide competition) See our DEMO
	Locating and Tracking of Facial Landmarks Using Gabor Wavelet Networks (with Roberto Cesar) International Conference on Advances in Pattern Recognition (ICAPR'01), Rio de Janeiro, Brazil, 2001 See also the related PROJECT PAGE
	A Project for Face Recognition from Video Sequences Using GWN and Eigenfeature Selection (with Teofilo Campos and Roberto Cesar) Workshop on Artificial Intelligence and Computer Vision (WAICV'2000), Atibaia, Brazil, 2000
	Improved Face versus Non-Face Discrimination Using Fourier Descriptors through Feature Selection (with Teofilo Campos and Roberto Cesar) IEEE Brazilian Symposium on Computer Graphics and Image Processing (Sibgrapi'2000), Gramado, Brazil, 2000
	Detection and Tracking of Facial Features in Video Sequences (with Teofilo Campos and Roberto Cesar) Lecture Notes on Artificial Intelligence, vol. 1793, pp. 127-135. Proceedings of MICAI-2000, Acapulco, Mexico, 2000. Springer-Verlag press
	Eigenfaces versus Eigeneyes: First Steps Towards Performance Assessment of Representations for Face Recognition (with Teofilo Campos and Roberto Cesar) Lecture Notes on Artificial Intelligence, vol. 1793, pp. 193-201. Proceedings of MICAI-2000, Acapulco, Mexico, 2000. Springer-Verlag press.