Eevee: Transforming Images to Bridge High-level and Low-level Understanding

<a href="https://embed.stanford.edu/iframe/?url=https%3A%2F%2Fpurl.stanford.edu%2Fdb907xw8028" class="su-underline">Show Content</a>

Abstract/Contents

Abstract: Images are central to communication, documentation, and a general understanding of the human world, and because their power so often arises from intangible subjectivities that currently elude computers, it is ever so vital that systems understand the world in the same way we do. People understand and think about images at a high level, but they’re currently constrained to work with images via low level tools and edit operations. We introduce a photo-editing system, Eevee, that can take a high level goal from a user and automatically translate it into a series of low level image edits, a step towards empowering people to interact with images at a higher semantic level. By situating our task within the domain of humorous memetic edits, building upon the notion of visual analogies to drive domain-shifting edits, and utilizing the objective knowledge afforded to us by modern computer vision technology (objects, their segmentations, their relationships, and their attributes), we produce a focused editing method that can link low level image syntax to high level image semantics.

Type of resource	text
Date created	June 4, 2018

Author	Lam, Michelle
Degree granting institution	Stanford University, Department of Computer Science
Primary advisor	Bernstein, Michael
Advisor	Krishna, Ranjay

Subject	Computer Science
Subject	Stanford
Subject	Stanford Engineering
Subject	Stanford School of Engineering
Subject	Human-computer interaction
Subject	HCI
Subject	Image editing
Subject	Photo editing
Subject	Graphics
Subject	Computer Vision
Genre	Thesis

Location	https://purl.stanford.edu/db907xw8028

Use and reproduction: User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
License: This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).

Preferred Citation: Lam, Michelle. (2018). Eevee: Transforming Images to Bridge High-level and Low-level Understanding. Stanford Digital Repository. Available at: https://purl.stanford.edu/db907xw8028

Undergraduate Theses, School of Engineering

Loading usage metrics...