Computer Science

Prof. Kwan-Liu Ma Awarded NSF Grant

Award ID: 1528203
NSF Program: INFO INTEGRATION & INFORMATICS
Title: Technologies for Creating Explanatory and Exploratory Animations from Scientific Data
PI: Kwan-Liu Ma
Period: Oct 1, 2015 – September 30, 2018
Amount: $499,996.00

Abstract:

Animation is a powerful, expressive medium for visual explanations and for telling stories with data. Scientists make extensive use of animations to explain their findings and to illustrate complex phenomena. By presenting time as time, animation is one of the most natural ways to illustrate how objects evolve and interact, and how they change in shape, size, position, and spatial relationship to other objects over time. Both commercial and open-source visualization tools offer a wealth of visualization techniques, enabling scientists to explore their data and to generate individual images to capture key aspects of the subject under study. However, most visualization packages include very limited support for creating explanatory animations. As a result, scientists who wish to use animations to illustrate their findings must spend considerable amounts of time learning how to produce animations, often using external software packages, or turn to professional animators or production specialists for assistance. This research aims to develop adequate support for composing animation content and constructing scientific video narratives, and also extend explanatory animation to exploratory animation and study its usability. This project will thus have a significant impact on both the visualization researchers and users in a range of domains, including education. The new concepts introduced in this project will inspire others to also develop similar and even better support for scientific storytelling using visualization. More users will benefit from such advanced visualization technologies leading to high productivity in their work, or support educational and outreach activities. The research team will continue collaboration with a science museum to seek the opportunities to convert explanatory/exploratory animations into interactive exhibits. Research will be integrated into teaching, in the form of special topic courses, the establishment of internships with industry and national laboratories, and the introduction of visualization technology to students from other disciplines. The project will provide an environment for research training for graduate and undergraduate students.

This research will introduce key technologies that can greatly increase scientists’ ability to make visualization animations and video narratives for storytelling. To facilitate scientific narrations using animations, this project will design a semi-automatic animation generation system tightly coupled with the interactive data exploration and visualization process. That is to make the process of animation and storytelling a first class citizen within exploratory data visualization tools. This will allow the scientists to focus on gaining insight from their data, and the visualization tools should assist them in assembling findings together into a coherent story for presentation. This project will design methods to choose views, camera paths, lighting, transitions, etc. for users. Methods for users to interact with animations will also be designed, instead of passively watching, to achieve new levels of inspection and apprehension. Explorable images, a powerful and novel concept introduced for realizing exploratory animation, enables multidimensional data exploration using a medium comparable to a video in terms of compactness and simplicity. The task of realizing these novel concepts and designs, and integrating them into scientists’ workflows and tools, will be challenging. This research will conduct extensive evaluation of the animation support, with the participation of domain scientists who are prospective users of the new technology. The lessons learned in this project may establish guidelines for the effective use of animation in explaining complex phenomena, and suggest a new framework for next-generation visualization systems. The project web site (http://vis.cs.ucdavis.edu/NSF/IIS1528203) will provide access to research results, including data and prototype software.

border