[robotics-worldwide] 2 Postdocs in Corso's Group at Buffalo on Robot Vision for Navigation and Interaction

Jason J. Corso jcorso at acm.org
Thu Jan 24 03:31:03 PST 2013


Applications are invited for two postdoctoral research associate
positions in Professor Jason Corso’s research group in Computer
Science and Engineering at SUNY at Buffalo.  The research area is
video understanding and search for robot navigation, interaction and
understanding.



Post-Doc 1: Joint Segmentation and Recognition of Activity in Video

The goal of the research project is to study the role of segmentation
in high-level video understanding.  Specifically, the project will
enable advanced activity recognition and video-to-text summarization
through joint activity recognition and segmentation.  The work will
study both surveillance footage and so-called videos-in-the-wild; it
will emphasize both event-level models as well as specific
shot-activity-recognition models.  The research will involve
supervoxel video over-segmentation, generatively and discriminatively
trained graphical models, and will emphasize human-specific actions.

The successful candidate will join this major effort involving
multiple senior investigators and 6 graduate students assistants and
play a critical role in making key research contributions. The
position is open to all US and non-US citizens who hold a Ph.D. in
computer science, applied mathematics/statistics, electrical
engineering, or equivalent, have a strong publishing record in
top-notch conferences and journals, have a strong understanding of
computer vision, machine learning, and knowledge
representation/ontologies, are experienced programmer beyond Matlab,
and have experience with UNIX platforms.


Post-Doc 2: Human-In-The-Loop Methods for Clustering and Metric
Learning for Video Search

The goal of the research project is to study human-in-the-loop and
active methods for clustering and metric learning in large-scale video
databases to facilitate search and retrieval of both set phenomena
(objects, actions, events) and ad hoc criteria set interactively by a
human user.  The work will investigate various active clustering and
semi-supervised metric learning methodologies with an emphasis on
novel methodologies.  More information about this project is available
at http://www.cse.buffalo.edu/~jcorso/r/ace/index.html

The successful candidate will join the PI and two graduate students
and play a critical role in making key research contributions. The
position is open to all US and non-US citizens who hold a Ph.D. in
computer science, applied mathematics/statistics, electrical
engineering, or equivalent, have a strong publishing record in
top-notch conferences and journals, have a strong understanding of
computer vision, machine learning, and knowledge
representation/ontologies, are experienced programmer beyond Matlab,
and have experience with UNIX platforms.


Contact and General Information:

Jason Corso  jcorso at buffalo.edu
http://www.cse.buffalo.edu/~jcorso


The candidates will have the opportunity to mentor graduate research assistants.

Salaries are competitive.

Dr. Corso runs a research group of 1 Research Scientist, 1 Post-doc,
12 Ph.D. Students, 3 Masters students and 3 undergraduates.  He is PI
on major externally funded research projects all in the area of
high-level computer vision, data mining and machine learning.  His lab
is 1250+ square feet with ample computing resources, including an
in-house Linux cluster, a 15TB SAN, and connectivity to the 2000+ node
Center for Computational Research in downtown Buffalo.  For both
openings, inquiries by e-mail (jcorso at buffalo.edu) are welcome before
a formal application.  An application requires a cv,
reprints/preprints, and the names of 3 references.


--
Assistant Professor
Computer Science and Engineering
SUNY at Buffalo

(e) jcorso at buffalo.edu
(w) http://www.cse.buffalo.edu/~jcorso
(o) 716-645-4754
(c) 716-208-3671
(f) 716-645-3464
(t) @ProfJasonCorso


More information about the robotics-worldwide mailing list