Dissertation/Thesis Abstract

Compositionality in vision and language
by Narayanaswamy, Siddharth, Ph.D., Purdue University, 2014, 119; 3636477
Abstract (Summary)

Compositionality can be found almost everywhere one looks. It is manifest in as diverse a range of entities as objects around us, the languages we use, and even our actions and interactions with the world. My work involves exploring and exploiting the general nature of such compositionality, often across multiple modalities such as vision and language, to solve deep and complex problems in perception. I demonstrate such ability in a variety of domains including part-based structures, board games, and activity recognition. I also show evidence for a particular kind of compositionality in how the brain perceives the world, lending further credence to the ubiquity and utility of compositionality.

Indexing (document details)
Advisor: Siskind, Jeffrey M.
Commitee: Davis, Larry S., Givan, Robert L., Wilbur, Ronnie B.
School: Purdue University
Department: Electrical and Computer Engineering
School Location: United States -- Indiana
Source: DAI-B 76/02(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Robotics, Artificial intelligence, Computer science
Keywords: Grounding language in vision, Language, Robotics, Vision
Publication Number: 3636477
ISBN: 9781321181203
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest