Eugene Levin, Michigan Tech University, Houghton MI USA
Background and Motivation:
In today’s world, the challenges related to efficient decision support during anthropogenic and natural environmental accidents, such as floods, hurricanes, earthquakes, oil spills, terrorist attacks, and others, are global by their nature. It requires the ability to collaborate on an international scale to prepare an efficient response. One of the promising fields of research is applying human-centric geospatial technologies as a way to utilize vast amounts of geospatial imaging data (aerial, satellite, SAR, LIDAR). The ultimate goal of cognitive geospatial technology research and development is establishing an interactive geospatial environment optimizing decision support workflow, making it more efficient, and accelerating productivity by producing automatic reactions to an analyst’s attention, emotions, and voice commands. Successful solution of the mentioned above problem requires interdisciplinary collaborations of scientists and engineers of multiple research fields including: geospatial sciences, computer sciences, cognitive science, neurosciences, knowledge engineering and others. Current seminar is devoted to present review of related developments worldwide and provide participant with profound understanding of the background and possible ways to achieve optimal human-computer symbiosis in geospatial environments domains.
Who should attend?
This seminar is targeted to undergraduate, graduate, PhD students and scientists who are interested in deploying efficiently human-computer interactions for their current and future research. There are no strict pre-requisites for the course, but previous accomplished on at least introductory level courses of Photogrammetry, GIS and Image interpretation will be very useful.
Seminar structure:
Seminar consists of two parts: 1) Theoretical background and review of cognitive geospatial research and related technology; 2) Open discussion, practical experiments and work on incorporation of cognitive workflow in participant-specific application scenarios.
Seminar agenda:
Part 1: Theory and review (morning)
1. Cognitive Sciences background: theory of human visual information perception
2. Research in perception of geospatial visual information — review.
3. Review in state-of-the-art eye-tracking and EEG technologies. Critical analysis of problems associated with eye-tracking technology — frequency, accuracy, calibration
4. Research in human-computer symbiosis as geospatial and neuroscience technology fusion:
a. Visual analytics and quality control of geospatial data;
b. Gaze-contingent display systems in Cognitive GIS;
c. Optimization of human-computer interactions for geospatial image interpretation environments;
d. Background in geospatial data mining knowledge engineering in application to geospatial imaging domain. Gaze-driven knowledge extraction and expertise transfer in geospatial image analysis;
e. Theory of human stereopsis: 3D measurements based on eye-trajectory measurements;
f. Perception of the bi-sensor/bi-temporal geospatial imagery for rapid data fusion and measurements;
g. Application scenarios of human-centric geospatial technologies.
Part 2: (Practice and discussion):
Experiment on eye-tracker calibration and sample data-sets recording.
Processing of results
Open discussion of student-specific application scenarios and workflows