In August 2010, I started as a software engineer with
Google in
Mountain View, CA.
research interests
I'm interested in artificial intelligence, machine learning, and
computer vision.
For my thesis, I'm working on using probabilistic geometric grammars
to represent and recognize classes of objects and scenes in
images.
While an undergrad at Stanford, I worked with Nils Nilsson on hierarchical relational
reinforcement learning, and with Pat Langley and Dan Shapiro on
hierarchical reactive cognitive architectures.
papers and presentations
Margaret Aycinena Lippow. "Weighted Geometric Grammars for Object
Detection in Context." Ph.D. thesis, Massachusetts Institute of
Technology, June
2010. pdf
Meg Aycinena Lippow, Leslie Pack Kaelbling, and Tomas
Lozano-Perez. "Learning Grammatical Models for Object Recognition."
In Proc. Daghstuhl Seminar on Logic and Probability for Scene
Interpretation, February 2008. urlpdf(short version of the following tech report)
Meg Aycinena, Leslie Pack Kaelbling, and Tomas
Lozano-Perez. "Learning Grammatical Models for Object Recognition."
MIT CSAIL Technical Report MIT-CSAIL-TR-2008-011, February
2008. pdfps
Meg Aycinena, Leslie Pack Kaelbling, and Tomas
Lozano-Perez. "Learning Grammatical Models for 2D Object Recognition."
Presented at the NIPS-2007 Workshop on the Grammar of Vision:
Probabilistic Grammar-Based Models for Visual Scene Understanding and
Object Categorization, December 2007. pptpdf
Margaret A. Aycinena. "Probabilistic Geometric Grammars for
Object Recognition." S.M. thesis, Massachusetts Institute of
Technology, August
2005. pspdf
Pat Langley, Daniel Shapiro, Meg Aycinena, and Michael
Siliski. "A value-driven architecture for intelligent behavior." In
Proc. IJCAI 2003 Workshop on Cognitive Modeling of Agents and
Multi-Agent Interactions, pp. 10-18, 2003. ps
"Aycinena" is pronounced phonetically, according to Spanish
pronunciation rules: "I-see-NAY-nah". The origin of the name is
Spanish Basque. "Lippow" is pronounced "LIPP-oh".