Session 1Q

Information Technology: Applications and Implications

1:00 PM to 2:30 PM | Moderated by Joseph Janes


City Tours
Presenter
  • Benjamin Andrew (Ben) Self, Junior, Computer Science Mary Gates Scholar
Mentors
  • Brian Curless,
  • Steven Seitz,
  • Avanish Kushal,
  • Yasutaka Furukawa,
Session
  • 1:00 PM to 2:30 PM

City Toursclose

Online photo-sharing sites such as Flickr contain vast amounts of freely available photos taken at sites all over the world. However, apart from keyword tagging, these photo collections do not have any structure and can only be viewed as simple slideshows. Our project focuses on using these photo collections to automatically generate visually pleasing tours of the sites. These tours consist of transitioning across a sequence of coherent images in 3D. We begin by reconstructing the geometry of the scene in the form of a 3D point cloud. We then use this data to generate a local model of the scene from the viewpoint of each camera. Next, we plan a path as a sequence of transitions between cameras through the scene which visits areas determined to be interesting. We rate the visual quality of each transition in order to select ones which are most visually pleasing. Finally, we present the tour by moving a virtual camera along the selected path and blending between the models of the two closest photos. During this presentation, we weight the contribution of each model based on how confident we are that the model is an accurate reconstruction.


Student Remote Document Collaboration
Presenters
  • Chris Mathews, Junior, Informatics
  • Genevieve Angela (Angela) Whittemore, Senior, Informatics (Human-Computer Interaction)
  • Charlie Tran, Senior, Informatics, Business Administration (Finance)
Mentor
  • Scott Barker,
Session
  • 1:00 PM to 2:30 PM

Student Remote Document Collaborationclose

In higher education, group projects that require students to create a single document have become very prevalent. Such projects bring both the benefits and challenges of collaboration. While a good deal of research and design work has been focused on supporting collaborative projects in the work environment, little research has examined how collaborative tools could improve project collaboration among students. Students in higher education can have different challenges than co-workers in a business environment, especially in terms of time allocation, scheduling, and how tasks are assigned to create documents. This research investigated if and how students remotely collaborate on documents, what collaboration tools they use, and how existing tools might be improved to better meet student needs in a remote collaboration environment. Initial in-depth interviews with individual students (N=10) were employed to gather qualitative data on students’ current habits. A subsequent survey to a larger sample (N=119) further explored student preferences and frustrations, and student focus groups (2 groups) explored design concepts to alleviate some of these frustrations. We found that students encountered group projects very often, predominately utilized two specific tools (i.e. email and Google Docs) in their collaborations, and were far from satisfied with their collaboration experiences. Students also preferred a simple interface for managing changes made to a document by teammates, a finding we discovered in testing design prototypes built expressly to address student frustrations identified in the interviews and surveys. Our research examined student habits, practices and frustrations in student group collaboration, with findings that will hopefully lead to fewer frustrations and have implications for the future of how students work together to achieve singular objectives in a learning environment.


Greater Access to Technologies in the Classroom Through Smarter Purchasing and Structural Decisions
Presenter
  • Victor H. Cantero, Senior, Spanish, Informatics Zesbaugh Scholar
Mentors
  • Tom Halverson,
  • Jenee Myers Twitchell,
Session
  • 1:00 PM to 2:30 PM

Greater Access to Technologies in the Classroom Through Smarter Purchasing and Structural Decisionsclose

As our world becomes modernized, and as communication, textbooks, collaboration between teachers, and even teaching in itself are digitized, one can no longer ignore the fact that access to technology in all forms is a necessity. Unfortunately, due to the current funding systems that schools are a part of, where property taxes make up the majority of a school’s income, many low-income schools cannot provide the technology or knowledgeable teachers to prepare students to succeed in a technology-driven job market. This study aims to analyze the Seattle School District’s purchasing methodology for classroom technology, as well as the overall use of current technologies: who is in charge of purchasing equipment, where the budget comes from, its mode of operation (centralized or decentralized), criteria and statistics of software/hardware updates, use of software/hardware in the classroom, percent of software features within technology actually used, among other aspects. The sole purpose is to find alternative ways of delivering the same level of technology access to low-income schools, or higher if possible, while maintaining or reducing the overall cost. This is done by looking at alternatives within the open-source realm, ranging from entire operating systems, to every day word processing tools, with the hope of eliminating the need for licensing fees that the Seattle Public School District must purchases on a regular basis in order to use software and receive services from companies like Microsoft. Finding more efficient ways to deliver quality technology access to all students would help Seattle Public Schools make better use of public funds to educate and introduce students to new technologies. As our nation experiences a funding crisis in education, these alternative ways of providing technology and its access will be relevant to the entire country, regardless of the district size or location.


TouchStone: Architectural Implications of a Multi-Touch Tabletop
Presenter
  • Imran Hasan (Imran) Peerbhai, Senior, Economics
Mentor
  • Brian Johnson,
Session
  • 1:00 PM to 2:30 PM

TouchStone: Architectural Implications of a Multi-Touch Tabletopclose

The widespread adoption of touch-screen devices and their increasing size, combined with the preference for “informal” interaction mechanisms (pinching, flicking, etc) on these devices, promises to enable a new era of “intuitive” finger-painting applications using sketch-like interfaces. In our own investigations of plan-sketching interfaces for casual users, one question that arose concerns the best way to support a user’s sense of scale and proportion in different drawing actions. Investigating this question, we utilized a DiamondTouch table and a simple floor-plan layout program (TouchStone) instrumented to record user interaction and editing actions. Our initial goal is to investigate how well users can draw “rooms” to specific dimensions given a variety of alternative scale cues. Future investigations will seek to learn and use individual layout preferences in feedback and automated assessment of algorithmically generated plans, examine how users respond to site information presented in the drawing environment, and investigate multi-user interaction and negotiation behaviors in an environment where no user is privileged by access to mouse or keyboard.


Enhancing Accessibility of Tactile Images through Audio
Presenter
  • Josh Scotland, Senior, Computer Science, Business Administration Mary Gates Scholar, NASA Space Grant Scholar
Mentor
  • Richard Ladner,
Session
  • 1:00 PM to 2:30 PM

Enhancing Accessibility of Tactile Images through Audioclose

Tactile graphics are embossed, nonvisual images designed for blind students. Image information is conveyed through touch and textual information is conveyed through the use of Braille. The research conducted for this project was aimed at making tactile graphics even more accessible to blind students than they currently are. This goal was achieved by using audio to express information rather than Braille because audio has two large accessibility benefits over Braille. First of all, Braille can clutter an image since it has to be a minimum of 29 point font, which is almost three times the size of typical Roman characters. The second accessibility issue of Braille is that less than ten percent of the blind student population in the United States uses it as a primary reading medium. Therefore, using audio instead of Braille can greatly enhance the accessibility of tactile images. There are two ways to add audio to a tactile image. One way is through a dot pattern printed on the page. A digital pen with a speaker can detect these dot patterns and play a unique audio file corresponding to the pattern area. Another way to add audio to a tactile image is to use barcodes. A smart phone application can scan these barcodes and play the text aloud. During my research, I created a digital pen application that identifies certain dot patterns and plays audio corresponding to these patterns. I also created a workflow to allow tactile graphic designers to quickly create audio tactile images. By using audio instead of Braille in a tactile image, we can substantially increase the accessibility of these images and give blind students more opportunities to succeed in their technical classes.


The University of Washington is committed to providing access and accommodation in its services, programs, and activities. To make a request connected to a disability or health condition contact the Office of Undergraduate Research at undergradresearch@uw.edu or the Disability Services Office at least ten days in advance.