About

AssembleMe is an information science blog written by Julius Schorzman that frequently sways off-topic.

Julius is the CEO of the Google Ventures backed company DailyCred. DailyCred makes working with OAuth super duper simple.

To view some of my old projects, visit Shopobot or CodeCodex.

You can follow me on Twitter if you really want to @schorzman.

Search
Contact Me
This form does not yet contain any fields.
    « The Year In Ideas | Main | Red States Ain't Abstaining »
    Wednesday
    Dec082004

    Interesting Seminars

    INFO SCIENCE: There are two interesting talks going on tonight at Intel Research Seattle. I won't be able to make either, but I thought I would pass the info along.



    Mary Czerwinski from the MSR will be speaking on Information Visualization and Large Display Research. Wednesday, December 8th, 4-5pm



    Abstract. Our early user studies documenting the increased productivity gained through the use of large displays allowed us to observe quite quickly that Windows and current applications do not scale well when vast amounts of screen real estate are available. Our group therefore set about iteratively designing software tools, based on real user problems, to support large-scale navigation and interaction. To ensure our software user interfaces provided value, we ran user studies against existing features and performed user-centered design. This talk will provide an overview of the prototypes we designed, and our methodology. In addition, I will discuss a few areas of long-term basic research on information visualization and interaction, and our attempts to scale the user experience across the spectrum of large and smaller displays.



    Bio. Mary Czerwinski is a Senior Researcher and Manager of the Visualization and Interaction Research group at Microsoft Research. The group is responsible for studying and designing advanced technology and interaction techniques that leverage human capabilities across a wide variety of input and output channels. Mary'' primary research areas include spatial cognition, information visualization and task switching. Mary has been an affiliate assistant professor at the Department of Psychology, University of Washington since 1996. She has also held positions at Compaq Computer Corporation, Rice University, Lockheed Engineering and Sciences Corporation, and Bell Communications Research. She received a Ph.D. in cognitive psychology from Indiana University in Bloomington. Mary is active in the field of Human-Computer Interaction, publishing and participating in a wide number of conferences, professional venues and journals.



    Erik Sudderth from MIT will be speaking on Visual Hand Tracking Using Nonparametric Belief Propagation. Wednesday, December 8th, 10:00 – 11:00pm



    Abstract. Probabilistic graphical models provide a powerful general framework for formulating and solving learning and inference problems. However, practical applications of graphical models in fields such as computer vision are often hampered by high-dimensional variables and non-linear relationships. In this talk, we describe nonparametric belief propagation (NBP), a stochastic algorithm which propagates sample-based approximations to true, continuous likelihoods. NBP effectively extends particle filters to the more general graphs rising in problems with spatial or hierarchical structure.



    We use NBP to visually track a 3D geometric hand model from image sequences, a challenging problem with applications in human-computer interfaces, motion capture, and scene understanding. Our tracker is based on a local representation chosen to allow information about each finger's location to guide the estimates of neighboring fingers. We show that the kinematic, structural, and temporal constraints underlying the tracking problem are naturally described by a graphical model. In addition, by introducing binary hidden variables describing the occlusion state of each pixel, NBP is able to properly reason about finger self-occlusions in a distributed fashion. Our results show that NBP may be used to refine inaccurate model initializations, as well as track hand motion through extended image sequences.



    Bio. Erik Sudderth is a doctoral student in the department of electrical engineering and computer science at the Massachusetts Institute of Technology, where he received the M.S. degree in 2002.He received the B.S. degree in electrical engineering from the University of California at San Diego in 1999. His research interests include statistical modeling and machine learning, and in particular the application of graphical models to problems in computer vision.

    Reader Comments

    There are no comments for this journal entry. To create a new comment, use the form below.

    PostPost a New Comment

    Enter your information below to add a new comment.

    My response is on my own website »
    Author Email (optional):
    Author URL (optional):
    Post:
     
    Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>