Lexical and Grammatical Inference

AAAI Conferences

Children are facile at both discovering word boundaries and using those words to build higher-level structures in tandem. Current research treats lexical acquisition and grammar induction as two distinct tasks; doing so has led to unreasonable assumptions. State-ofthe-art unsupervised results presuppose a perfectly segmented, noise-free lexicon, while largely ignoring how the lexicon is used. This paper combines both tasks in a novel framework for bootstrapping lexical acquisition and grammar induction.


Object Recognition Based on Visual Grammars and Bayesian Networks

AAAI Conferences

A novel proposal for object recognition based on relational grammars and Bayesian Networks is presented. Based on this grammar an object is represented as a hierarchy of features and spatial relations. This representation is transformed to a Bayesian network structure which parameters are learned from examples. Thus, recognition is based on probabilistic inference in the Bayesian network representation. Preliminary results in modeling natural objects are presented.


Learning in the Lexical-Grammatical Interface

AAAI Conferences

Children are facile at both discovering word boundaries and using those words to build higher-level structures in tandem. Current research treats lexical acquisition and grammar induction as two distinct tasks. Doing so has led to unreasonable assumptions. Existing work in grammar induction presupposes a perfectly segmented, noise-free lexicon, while lexical learning approaches largely ignore how the lexicon is used. This paper combines both tasks in a novel framework for bootstrapping lexical acquisition and grammar induction.


Submodular Field Grammars: Representation, Inference, and Application to Image Parsing

Neural Information Processing Systems

Natural scenes contain many layers of part-subpart structure, and distributions over them are thus naturally represented by stochastic image grammars, with one production per decomposition of a part. Unfortunately, in contrast to language grammars, where the number of possible split points for a production $A \rightarrow BC$ is linear in the length of $A$, in an image there are an exponential number of ways to split a region into subregions. This makes parsing intractable and requires image grammars to be severely restricted in practice, for example by allowing only rectangular regions. In this paper, we address this problem by associating with each production a submodular Markov random field whose labels are the subparts and whose labeling segments the current object into these subparts. We call the result a submodular field grammar (SFG). Finding the MAP split of a region into subregions is now tractable, and by exploiting this we develop an efficient approximate algorithm for MAP parsing of images with SFGs. Empirically, we present promising improvements in accuracy when using SFGs for scene understanding, and show exponential improvements in inference time compared to traditional methods, while returning comparable minima.


Submodular Field Grammars: Representation, Inference, and Application to Image Parsing

Neural Information Processing Systems

Natural scenes contain many layers of part-subpart structure, and distributions over them are thus naturally represented by stochastic image grammars, with one production per decomposition of a part. Unfortunately, in contrast to language grammars, where the number of possible split points for a production $A \rightarrow BC$ is linear in the length of $A$, in an image there are an exponential number of ways to split a region into subregions. This makes parsing intractable and requires image grammars to be severely restricted in practice, for example by allowing only rectangular regions. In this paper, we address this problem by associating with each production a submodular Markov random field whose labels are the subparts and whose labeling segments the current object into these subparts. We call the result a submodular field grammar (SFG). Finding the MAP split of a region into subregions is now tractable, and by exploiting this we develop an efficient approximate algorithm for MAP parsing of images with SFGs. Empirically, we present promising improvements in accuracy when using SFGs for scene understanding, and show exponential improvements in inference time compared to traditional methods, while returning comparable minima.