|
NSF IUSE:
Improving undergraduate student critical thinking and ability to solve environmental problems with fossil records through FossilSketch application. PI Stepanova. CoPIs Belanger, Anwar, Hammond 2024-02-01–2027-01-31, $749,811 Abstract: |
|
|
NSF HSI:
HSI Implementation and Evaluation Project: SedimentSketch, teaching tool in and beyond the sedimentology classroom to provide equitable and inclusive learning for Hispanic students PI Hammond CoPIS Juan Carlos Laya, Anna Stepanova, Carlos Alvarez Zarikian, Saira Anwar. 2024-09-01–2026-07-31, $392,829 Abstract: |
|
|
NSF IUSE:
Collaborative Research: Fostering Engineering Creativity and Communication through Immediate, Personalized Feedback on 2D-Perspective Drawing TAMU (Lead) PI: Tracy Hammond, CoPI: Vinayak Krishnamurthy, Purdue PI: Kerrie Douglas, Georgia Tech PI: Julie Linsey, CoPI: Wayne Li, San Jose State PI: Vimal Viswanathan 2020-07-01–2024-06-30, $1,500,000 Abstract: This project aims to serve the national interest in excellent undergraduate engineering education by improving students' ability to draw representations of structures and systems. Free-hand drawing is a crucial skill across engineering, especially for its ability to reduce complex systems to simplified, accurate diagrams. Such diagrams aid in idea generation, visualization of systems, and discussions between clients and engineers. When coupled with appropriate timely feedback, drawing can also be an effective learning tool for improving students' visual communication skills and creativity. This project will use a computer application called SketchTivity to teach engineering students how to draw, and examine the impact of drawing instruction on student learning. SketchTivity is an intelligent tutoring system that provides real-time feedback on 2D drawings that students make on a screen instead of on paper. The application, which was developed by the research team, provides each student with iterative, real-time, personalized feedback on the drawing, promoting improvements and facilitating learning. Free-hand drawing and its associated benefits were inadvertently removed from the engineering curricula when educators transitioned from hand drafting to Computer-Aided-Design. This transition has resulted in the lack of student and faculty proficiency in freehand drawing. This project will support restoration of 2D-perspective drawing to engineering curricula. Since drawing is a skill that is also relevant to other STEM fields, this work is likely to have broad relevance in undergraduate STEM education. The project will enable the distribution of SketchTivity software to approximately 5,000 diverse undergraduate and graduate students at four partnering institutions. These students include a large percentage of women and students from other groups that are not equitably represented in STEM fields. The project has the potential to produce significant new knowledge about drawing-based artificial intelligence (AI) tools. It can also increase understanding about the effects of feedback and reflective prompts on drawing skills, and on learning and creativity. The project will compare AI-based assessments to assessments conducted by humans, thus providing additional information about development of intelligent tutoring systems. The impact of drawing ability on students' creativity and spatial reasoning skills will be investigated, along with the transferability of drawing skills to other courses in engineering. A mixed-methods approach that includes surveys, validated assessments for engineering design creativity, and drawing quizzes, is proposed to assess project outcomes. The NSF IUSE: EHR Program supports research and development projects to improve the effectiveness of STEM education for all students. Through the Engaged Student Learning track, the program supports the creation, exploration, and implementation of promising practices and tools. |
|
|
NSF Earth Sciences:
Collaborative Research: Community tools for automated paleoenvironmental interpretation from sedimentary field data Penn State (Lead) PI: Liz Hajek; TAMU PI:Ryan Ewing; Co-PI:Tracy Hammond; 2020-06-01–2023-05-30, $897,739 Abstract: Geological field data is essential for reconstructing historical conditions on Earth and Mars, finding and developing natural resources, and managing natural hazards. Sedimentary geology relies on a set of patterns in rock outcrops that provide information on where natural environments (like oceans, rivers, deserts, or lakes) existed in the past. This project will develop tools for automated interpretation of past environments from sedimentary outcrop data. Collaborative evaluation of outcrop datasets by geoscientists and computer scientists will yield new perspectives on how geological outcrop interpretation can be accomplished and provide a deeper understanding of the information required for automated interpretation of field data. Automating outcrop interpretation will broaden researchers' ability to rapidly mine existing and new digital datasets for information that will contribute new insights into the history of surface conditions on Earth and Mars. The tools developed as part of this project will be broadly disseminated and will integrate with existing digital platforms for field data. These tools will help facilitate collaborations among a broad range of geoscientists, including those who can't easily conduct fieldwork in remote locations. This project will strengthen workforce development by cross-training sedimentary geologists with advanced computer science skills and will provide examples of practical applications of machine-learning approaches for computer science students. To accomplish these goals, this project will leverage existing digital outcrop datasets and collect a new targeted dataset to explore automated approaches for extracting sedimentary features from outcrop image and surface-topography data. The project will focus on extracting sedimentary features that are critical for paleoenvironmental reconstruction, including types of cross bedding, and investigate how outcrop quality, scale, and orientation influence the recoverability of sedimentary features. Additionally, this project will explore the degree to which the three-dimensional orientation of sedimentary features can be extrapolated from automated outcrop observations. The tools developed in this project will be tested with structural geology and geomorphology outcrop datasets to evaluate how workflows developed as part of this work could aid a broader range of geoscience disciplines. Results will be packaged as an accessible online tool that can be used in combination with digital field data repositories. Widespread dissemination of this tool and broad community participation will strengthen the algorithm and improve the community resource over time. Educational materials suitable for undergraduate sedimentary geology courses will be developed and will help expand undergraduate geosciences students' exposure to computer science. |
|
|
NSF IUSE:
Developing a Digital Sketching Application That Delivers Personalized Feedback to Improve Student Learning and Engagement in Micropaleontology PI: Tracy Hammond; Co-PI: Christina Belanger, Anna Stepanova, Christine Stanley, Sara Raven 2020-01-01–2022-12-13, $300,000 Abstract: This project aims to serve the national interest by improving undergraduate geoscience education. It focuses on micropaleontology, which is the study of microscopic fossils in rocks, sand, and other geological samples. Information about the type and number of microfossils in a sample is useful for understanding the history of the earth. It also has applications in oil, engineering, mining, and other industries. This project will develop new teaching materials and an interactive software application to help students to learn how to identify microfossils. This application, called FossilSketch, will help students make accurate digital sketches of microfossils on touch-screen devices. Once a student has made an initial sketch of a microfossil using FossilSketch, the software will use an artificial intelligence algorithm to provide automated, instantaneous feedback on the sketch. By working to improve the accuracy of their sketches, students will develop the knowledge and observational skills needed to identify microfossils in real samples. This mastery will not only help students learn micropaleontology, but also enhance their ability to use micropaleontology in undergraduate research projects. Thus, the project has the potential to enhance geosciences education without increasing faculty workload. In addition, by supporting student success, it can also broaden participation of students in geosciences majors and careers. FossilSketch will be the first teaching tool in micropaleontology that combines pen-and-touch technology with principles of sketch recognition to train undergraduate students. The application will focus on Foraminifera and Ostracoda, the two taxonomic groups most actively used in geological and environmental studies. This project aims to examine the impact of this sketch-recognition based tutoring system on students' comprehension and retention of micropaleontological concepts. It will also measure the impact of the system on students' interest and self-efficacy in geosciences. The FossilSketch system will be web-based and platform independent, so that it can work on all devices and be accessible to a wide audience. Novel, microfossil-specific sketch recognition algorithms will be developed to accommodate the distinct morphological diversity of microfossils, support increased sketch-recognition accuracy and speed, and provide improved feedback mechanisms for students and instructors. The classroom use of FossilSketch will be assessed in several geoscience courses at Texas A&M University. Project outcomes and the FossilSketch software will be shared through the project website, professional development workshops, and publications. This project is supported by the NSF Improving Undergraduate STEM Education Program: Education and Human Resources Program, which supports research and development projects to improve the effectiveness of STEM education for all students. Through the Engaged Student Learning track, the program supports the creation, exploration, and implementation of promising practices and tools. |
|
|
NSF:
NSF DUE 1726306 Enhancing Visualization Skills and Conceptual Understanding Using a Drawing-Recognition Tutoring System for Engineering Students TAMU PIs: Tracy Hammond, Kristi Shryock, Stephanie Valentine; Georgia Tech PI: Julie Linsey; LaTourneau PI: Brian Caldwell; San Jose State PI: Vimal Viswanathan; Texas State PI: Kimberly Talley 2017-09-01–2022-08-31, $1,814,400 Abstract: Visual and spatial skills are important for scientific and engineering innovation. The ability to represent real systems through accurate yet simplified diagrams is a crucial skill for engineers. A growing concern among engineering educators is that students are losing both the skill of sketching and the ability to produce the free-body diagrams (FBDs) of real systems. These diagrams form the basis for various types of engineering analyses. To address this concern, investigators will redesign and test a cutting-edge educational technology for engineering concepts of statics and mechanics. The sketch-based technology developed at Texas A&M University, called Mechanix, enabled students to hand-draw FBDs, trusses, and other objects using digital ink and provided helpful feedback. The upgraded Mechanix software will include enhanced artificial intelligence (AI) to understand the sketches and provide immediate feedback to the student for individualized tutoring. Instructors will also receive real-time detailed information from the system so they can clarify misconceptions and guide students through problem solutions during classes. This free-hand sketch-based system will focus learning on the fundamental engineering concepts and not on how to use a software tool. These engineering concepts directly relate to a wide variety of designs including bridges, buildings, and trusses that are vital to the infrastructure of the nation's cities. The project will help prepare engineers with improved abilities to develop these designs that are essential in society. This project will aim to demonstrate the impact of the sketch-recognition based tutoring system on students' motivation and learning outcomes, both generally and among students of diverse backgrounds. The Mechanix system will be converted to an HTML5 format to work on all devices and expand its accessibility for institutions with various technological requirements. Additional AI algorithms will be developed to accommodate more types of statics problems, increased sketch-recognition accuracy and speed, and improved feedback mechanisms for instructors that merge performance information for the students in a class. The upgraded system will be studied in various engineering courses across five different universities, and introduced to over 2,500 students in engineering and related fields. The investigators will utilize controlled classroom experiments, digital data collection, pre/post concept testing, focus groups, and interviews to explore the external validity of Mechanix as a learning tool. Analysis of Covariance will be used to compare outcomes for students using Mechanix and students in control groups. Project outcomes and the Mechanix software will be shared through the project website, professional development workshops, and publications. |
|
|
NSF Smart and Connected Communities:
SCC-PG: Fostering Aging-in-Place and Autonomy in Elderly Persons through Intelligent Tracking PI: Tracy Hammond, CoPI Daniel Goldberg 2020-10-01–2021-09-30, $150,000 Abstract: As people age they begin to need assistance with activities of daily life (ADLs). Losing the ability to perform ADLs requires that these individuals receive care from family members, move to senior living communities or experience a significant drop in their quality of life. Given the projected rise in the number of older adults over the coming decades, the resources available to care for this population will be spread thin, potentially leaving some with lower quality of care or even without any form of care. One means of addressing this is developing systems that allow older adults to live independently while maintaining their quality of life for longer. Such systems would not only meet a well-documented desire to live independently, but also reduce the burden of care on caregivers and/or the need for more expensive forms of long-term care. This project will lay the groundwork for the development of a scalable gerontechnological system consisting of sensors and novel algorithms designed to recognize ADL performance in real-world settings. Through tracking of resident performance of ADLs we aim to facilitate more timely and relevant care and interventions. The project works with multiple senior living communities to develop a gerontechnological system that allows for a quantified daily health profile to enable meaningful care for those who wish to age-in-place. This planning grant attempts to answer the following research questions through interviews and observational studies with each collaborating senior living community and determination of preliminary technological details: (1) What are the specific needs of both caregivers and residents from a gerontechnological system built on the recognition of ADLs and resident location' (2) What are the requirements of the indoor positioning system (IPS)' (3) What are the privacy needs and concerns of the caretakers and residents and how can the system address them' The information gained from the planning grant will allow us to submit a multi-year Integrative Research Grant that will attempt to answer the following research questions: (1) Can common ADLs be accurately recognized using a combination of wearable devices and novel machine learning algorithms and methodologies' (2) How does the recognition of activity patterns, feedback to residents and caregivers, and individually-tailored interventions support and/or impact the lives of senior living community residents' (3) How does the recognition of activity patterns, feedback to residents and caregivers, and individually-tailored ADL support and/or impact the workload and ease the burden of care on the caregivers within the senior living facilities' Successful completion of this work has the potential to transform the current paradigm of elderly healthcare from reactive to proactive by monitoring and supporting the health of the elderly population. |
|
|
NSF Future of Work:
FW-HTF-P: A Socio-technical Approach to Help the HR Function of the Future: Identifying and Preventing Discriminatory Recruitment Practices in the Technology Industry PI: Tracy Hammond; CoPI: Joanna Lahey, Sherecce Fields 2020-08-15–2021-07-14, $150,000 Abstract: Despite greater awareness of the value of broadening participation in STEM fields and movements to do so, discrimination in hiring remains a problem. The high technology sector is particularly notable for the very low numbers of women, African American, and Hispanic American computer scientists and engineers. A key reason for the low participation is that implicit biases at different stages of recruitment perpetuate the inequitable status quo. To begin to solve this problem, discrimination in recruiting must be understood in detail. This work aims to work toward improving human resources (HR) practices by improving training that uses state-of-the-art technologies to help people avoid unconsciously falling into discriminatory behavior. A potential long-term societal benefit of this work would be more equitable hiring processes. Such an improvement could increase hiring of diverse teams, thus increasing creativity and innovation in the workplace. In addition, more diverse teams can help to reduce stereotype threat, since sterotypes break down as more people from underrepresented groups are included. Toward these goals, the research team plans to conduct an interdisciplinary workshop with industry representatives and experts from HR management, the social sciences, and computer science to identify sources of bias and propose solutions to address them. This workshop is expected to lay the foundation and the research agenda necessary to submit a full Future of Work at the Human-Technology Frontier proposal. By leveraging the expertise of HR management, psychology, and sociology professionals, in partnership with computer scientists and engineers, this convergent research team plans to gain a deep understanding of where and what kinds of bias occur in the recruitment process for high technology workers. Participants will explore potential interventions to identify or prevent bias through the use of future technologies such as virtual reality, augmented reality, and artificial intelligence-powered tools. The research team will attempt to answer the following questions through this project: What are the standard practices of the technology industry's recruitment process' How much do practices vary from company to company' In what phases of the recruitment process does bias occur, and what types of biases occur' What types of evidence are produced by biased recruiting practices' What technologies are useful for identifying and preventing bias' This project has been funded by the Future of Work at the Human-Technology Frontier cross-directorate program to facilitate convergent research to promote deeper basic understanding of the interdependent human-technology partnership to advance societal needs by advancing design of intelligent work technologies that operate in harmony with human workers. |
|
Texas A&M Triads for Transformation (T3):
Online Local Environment Hazards Education for Young Adults using a Social Media Platform PI: Courtney Thompson, CoPIs: Tracy Hammond and Daniel Goldberg 2019-01-01–2020-12-31, $36,025 Abstract: Natural hazards education has been found to reduce disaster impacts at individual and community levels. During disasters, people choose to act depending on how they perceive a hazard or risk. However, studies often omit the severity to which children and young adults experience disasters, including changes in risk perception. In addition, curriculum in most secondary schools does not cover local natural hazards or their impacts in sufficient depth. This study aimed to develop a formal child-centric natural hazard and disaster educational program while investigating how the program influences risk perceptions of local natural hazards. Local college students were the preliminary subjects to ensure the program’s quality and efficacy. |
|
|
NSF SMA:
SBE: 1560106 REU Site: Cyber-Health GIS – Multidisciplinary Research Experiences in Spatial Dynamics of Health Project PI: Daniel Goldberg, CoPI: Tracy Hammond 2016-03-01–2020-02-29, $344,674 Abstract: This project is funded from the Research Experiences for Undergraduates (REU) Sites program in the SBE Directorate. As such, it has both scientific and societal benefits, and it integrates research and education. This REU Site combines the fields of Computing, Geographic Information Science (GIS), and Health on the Texas A&M University (TAMU) campus in College Station, TX where undergraduate students from diverse backgrounds in Computing, GIS, and Health work together on collaborative research projects in a newly emerging field called Cyber-Enabled HealthGIS (Cyber-HealthGIS). Thirty REU students (ten each year) are engaged in research teams to promote discovery, teaching, and training through hands-on research and mentoring. Students are mentored and trained in the basics of research techniques, the responsible conduct of research, the need for diversity in research, and research designs and methods. The REU Site students collaboratively pose, execute, and evaluate research projects resulting in research advances in Cyber-HealthGIS. Through this approach, students learn research and problem-solving outside of their own discipline, and gain independence and confidence in their own ability to undertake research. This program will advance the new field of Cyber-HealthGIS by building theory, methods, and approaches which will lead to joint faculty-student publications in research journals, seminars, and conferences, and student presentations of their own research. This program fills a critical US workforce gap by creating a generation of students trained for and interested in research and scientific careers in Cyber-HealthGIS, a rapidly advancing field with the potential to improve human health and well-being. The student projects and example prototypes developed through this program will be made freely available to help foster innovation and development in the Cyber-HealthGIS industry. In this REU Site, the ideas from each student's academic/disciplinary domain is integrated to form a cohesive, achievable research goal under the umbrella of the core research themes of this project, which include (1) Outbreak surveillance through the combination of authoritative and social media data; (2) High-resolution chronic disease risk mapping with citizen-derived perceptions of community; and (3) Continuous time-enabled scalable outbreak planning. Students work closely with faculty mentors to pose the research question, develop testable hypotheses, obtain the necessary data, organize appropriate methods, engineer an approach, perform experiments, and undertake an evaluation of the results. This project will make freely available examples of prototype applications, thereby advancing the capabilities of Cyber-HealthGIS research through the development and release of free and open source (FOSS) code for the systems and example data sets used in the student research and experiments. |
|
|
NSF:
SBP: 1658758 Collaborative Research: Gender Discrimination in Hiring for STEM Graduates PI: Joanna Lahey & Gerianne Alexander, CoPI: Tracy Hammond 2017-03-01–2020-02-28, $465,523 Abstract: This interdisciplinary project will use cutting-edge technology to study the labor market for computer science graduates in Science, Technology, Engineering, and Mathematics (STEM) fields. Although women's share in STEM employment has been growing in non-computer science occupations, their share in computer science occupations has been declining since the 1990s. Because computer science occupations account for 50% of STEM workers, this decline is slowing the growth of women's share in STEM fields overall, and suggests significant untapped potential that could improve US productivity and competitiveness. One reason that women may not seek out or remain in computer science fields is that they are treated differently than men during the hiring process. This project uses a laboratory experiment in the field on first-line hiring managers to determine first if there is differential treatment of women in hiring recent computer science graduates. If there is such a difference, it will determine the characteristics of women who are more likely to be treated negatively as well as general characteristics of firms that are more likely to exhibit differential treatment. Resumes will be randomly generated to include different characteristics that, if their inclusion helps women more than men, will indicate potential reasons for this differential treatment. The experiment will also use eye-tracking to determine how recruiters visually process computer science resumes and whether or not there are differences between how they process male vs. female resumes. These combined results will help to differentiate between economic theories of discrimination, and will advance social science by increasing our theoretical understanding of when and how differential treatment occurs. Results from this study can be used to make recommendations to individuals applying for these positions and institutions which advise them, to employers who desire to hire the best candidates, and to policy makers who want to increase meritocratic hiring in STEM. The results will thus lead to a more diverse and competitive workforce, increasing the economic competitiveness of the U.S.< br>This project combines two cutting-edge methodologies, eye-tracking and resume-randomization, to study gender discrimination at the first stage of the STEM hiring process. It will determine if there is differential treatment by gender in how first-line hiring managers treat resumes, whether the treatment is similar or different along the applicant quality distribution, and if there are industry characteristics (ex. firm size, industry code) that would lead to higher or lower levels of differential treatment. Finally, this study will differentiate between different theories of statistical and taste-based discrimination. Technical recruiters in charge of first-line interview decisions will be solicited at university recruitment fairs and industry fairs to view and process hypothetical resumes for Computer Science majors. They will be asked to follow their standard hiring practice and to choose resumes to "move to the next stage." The resumes will then be redisplayed and participants will rate each resume and give the expected starting salary and position. While participants are viewing the resumes, their eye-movements will be tracked via an eye-tracking device. Following the resume rating exercise, they will answer a short demographic survey. Resumes with randomized inputs based on actual resumes will be created via a randomization program. Outcomes of interest include information on ratings, moving the resume to the next stage, position placement, salary ranges, time spent on individual resumes, time spent on and number of looks at specific parts of resumes. The coefficients and significance on the coefficient of gender determine whether or not there is differential treatment by gender, and if so, for which women and by what kinds of firms. Time spent on resumes by gender interacted with differential treatment findings provide information on use of heuristics in the decision-making process. Time spent viewing specific parts of the resume (areas of interest or AOI) and tracking the order that recruiters view parts of the resume provide insight into their decision-making processes. Gender interactions with randomized inputs that support or contradict stereotypes will be used to test employee taste-based discrimination and levels-based statistical discrimination. Position placement by gender will test customer taste-based discrimination. Comparing predicted outcomes with actual outcomes by gender of resume be used to test variance-based statistical discrimination. This project directly impacts the full participation of women in STEM and will (1) improve the well-being of individuals in society, (2) develop a diverse and competitive workforce and (3) increase economic competitiveness. Results from this study can be used to make recommendations to individuals applying for these positions and the institutions who advise them, to employers who desire to hire the best candidates, and to policy makers who want more women and minorities in STEM. The methodology will (4) promote future research on other hiring and discrimination questions. In addition, this project will (5) incorporate graduate and undergraduate students, involving them in cutting-edge research and providing them with a platform to undertake their own independent work. Graduate and undergraduate students will receive mentoring and research skills, increasing their attractiveness to employers and advanced degree programs. |
|
NIH:
NIH 1R01CA197761-01 Physical Activity Impacts of a Planned Activity-Friendly Community: The What, Where, When, and Why of Environmental Approaches to Obesity Prevention PI: Marcia Ory 2015-02-01–2020-01-31, $3,325,801 Abstract: Physical activity (PA) helps prevent obesity and reduce chronic conditions such as cancer, diabetes, and heart disease, and can be promoted through environmental/policy interventions at the population level. However, existing empirical knowledge on environment-PA relationships is primarily based on cross-sectional studies, which provide insufficient control of extraneous factors for investigating causal relationships. Further, little is known about how environmental factors affect spatial (where) and temporal (when) patterns of PA and the underlying mechanisms (why) of such impacts, including potential mediating effects of the psychosocial factors. The objective of this longitudinal, case-comparison study is to examine both short-term and long-term changes in PA after residents move to an activity-friendly community (AFC). It utilizes a unique and fleeting opportunity with ~3000 new homes being built in a large planned AFC over the next ~3 years. The focus is on those who are currently sedentary or insufficiently active and living in an environment lacking support for PA. Case participants (n~350) are those adults moving from non-AFCs to this AFC and not meeting the CDC guidelines for PA at pre-move baseline. Each case participant will be matched based on gender and age (±5 years) with a comparison participant who lives in his/her pre-move non-AFC, is also sedentary or insufficiently active, and is not planning to move for at least two years (this project's follow-up measurement period). The specific aims of this proposed study are to (1) examine the short-term and long-term changes in total PA levels (weekly minutes) and in spatial and temporal patterns of PA (proportion of PA taking place within the community, proportion of walking out of total PA, and level of PA integration into daily routines), after sedentary or insufficiently active individuals move from non-AFCs to an AFC; and (2) determine what built and natural environmental factors (e.g. density, land uses, sidewalks, trails/paths, parks, water features) lead to changes in PA among these populations, either directly or indirectly by affecting psychosocial factors related to PA. Using this timely opportuniy to gain longitudinal assessments for this natural experiment is of critical importance to advancing the status of knowledge on the intersection of health and place as it relates to promoting PA. The interdisciplinary research team has extensive experiences on this topic and in this study community through pilot work. At the conclusion of this study, we will have identified stronger evidence supporting the impact of an AFC on population-level behavior changes toward more physically active lifestyles (short-term goal) and toward lessening the burden of obesity throughout the nation (long-term goal). |
|
|
NSF IIS 1649126::
EAGER: Exploring Children's Use of Online Social Networks Using the KidGab Network PI: Tracy Hammond, CoPI: Cara Wallis, Stephanie Valentine 2016-07-01–2019-06-30, $191,545 Abstract: This proposal will support research into why children use social networks and how they influence each other using KidGab, a social network managed by the research team and designed for pre-teen Girl Scouts. Although pre-teens regularly use social networks, relatively little is known about how they (versus adults) behave online or how this affects their well-being because most popular social networks close pre-teens' accounts when detected, while purpose-built networks for pre-teens are heavily restricted in terms of both what users can do and what data is available for study. Building on their existing work with Girl Scout councils in Texas, the research team will develop new features and activities for KidGab and conduct outreach workshops with new councils. This will allow the team to study how children respond to different recruitment and motivating ('gamification') strategies, in particular, looking at the relative value of adult- versus peer-created content and collaborative versus individual activities in encouraging continued use. Through looking at how children create and adopt each other's drawings in visually-focused activities, the team will also develop novel methods for studying creativity, conformity, and influence in social networks. In addition to making progress on these specific questions, the team's long-term research goal is to build a large enough network and dataset that both they and other researchers can conduct future studies and analyses. More broadly, the team will create useful online content and design guidance for building social networks for pre-teens that support positive outcomes such as identity development and personal connection while reducing bad outcomes such as oversharing and cyberbullying. Preliminary work by the team shows that the regular release of new content is critical to retaining participation; thus, the first main component of the proposal is to develop ways for children in the network to generate and share content such as personality quizzes, and images that network members can use virtual currency to buy and display on their profiles. This will allow the team to (1) compare the uptake of adult-created versus peer-created content and their effects on encouraging long-term participation, (2) examine the kinds of content children prefer to generate and consume while generating a library of child-created content, and (3) study larger questions about identity creation and exploration. The second main component is to develop sketching-based activities in which participants are given a creativity task and their work is made visible through the network so that other participants can adopt ideas from it in their own work. The team will manually code key features of sketches generated for a given task and study their propagation using link-analysis algorithms such as PageRank and the Hubs and Authorities Algorithm, interpreting the degree to which a participant is a hub or authority as the likelihood that they are influenced by or influence others. By looking at a variety of specific tasks and variations in instructions that prime behaviors, as well as characteristics of participants, the team will develop insights into key drivers of influence in pre-teens' social networks. Further, the manually annotated sketches will provide training data for computer vision and machine learning algorithms for sketch analysis. The team will deploy these content creation mechanisms through events held with individual Scout councils geographically near those who have already participated in the network; such a strategy will best leverage the team's existing relationships with nearby councils while increasing the chance of recruiting dense sub-networks to encourage long-term retention. |
|
|
NSF IIS Cyberlearning 1441331:
EXP: Collaborative Research: PerSketchTivity—Empowering and Inspiring Creative, Competent, Communicative, and Effective Engineers through Perspective Sketching TAMU (Lead) PI: Tracy Hammond, CoPIs: Erin McTigue and Jeff Liew, Georgia Tech PI: Julie Linsey, CoPI: Wayne Li 2014-09-01–2019-06-30, $613,055 Abstract: This project examines whether technology can support learning to freehand sketch. Sketching has been demonstrated to play an important role in a number of domains, including engineering, and the ability to quickly sketch has been shown to improve creativity by making it easier for engineers to generate ideas and communicate them. This project will modify artificial intelligence tools that support recognizing sketches to directly help teach undergraduate engineers how to sketch well. Research studies will examine whether the tool helps students learn sketching skills, and importantly how it influences their spatial reasoning ability. Thus, if successful this research will not only create tools to allow people to learn to sketch better, but also will advance our understanding of how spatial reasoning and sketching are linked, and could eventually lead to more effective engineering education. The project proposes two interconnected strands of work: developing the software tool and conducting research studies in the context of undergraduate engineering courses. The software tool will use a heterogenous set of classifiers to help provide feedback to learners as they perform a sequence of sketching exercises on tablets. The design process will iterate on the tool to explore what types of feedback are most helpful and how different classifiers can be used to detect different levels of sketching skill. The program of research will include studying whether sketching training leads to advances in spatial reasoning skills, whether it affects design self-efficacy and attitudes towards sketching, transfer of spatial skillsets to design activities in other courses, and how sketching skills correlate to success on spatial reasoning tasks. In addition, through iterative development including user-centered design processes, design principles for sketching based tools will be derived. Data sources will include both qualitative and quantitative data such as pre- and post-test spatial reasoning tasks, structured interviews, surveys, and artifact analysis. Additionally, students (N=approximately 30-40) using the new tool in class will be compared to control cohorts of approximately 30 students who either use traditional engineering curricula (little free-hand sketching and some isometric drawing) and a sketching curriculum without the AI tool. |
|
ESRI:
SRLx (Sketch Recognition Lab eXperience 2017) PI: Tracy Hammond 2017-11-01–2018-10-31, $1,000 Abstract: SRLx, directed by Dr. Tracy Hammond, is a multi-day symposium highlighting state-of-the-art research on activity recognition (eye, body, and sketch motions), haptics, intelligent fabrics, app development, Computer Vision, AI, Big Data, Machine Learning, HCI, and UX Design. Presenters will include Sketch Recognition Lab members, alumni, and collaborators. Participants will learn about and discuss the implications of ground-breaking research in both academia and industry. |
|
|
Texas A&M College of Engineering AggiE-Challenge:
Developing a MOOCS Platform for Online Personalized Learning allowing Sketch Input PI: Tracy Hammond 2013-01-01–2018-05-31, $166,000 Abstract: AggiE_Challenge is designed to engage engineering undergraduate students with multidisciplinary team research projects related to engineering challenges facing our society. The grand challenges include the 14 Grand Challenges for Engineering (National Academy of Engineering), the 14 Grand Challenges for Global Health (Bill and Melinda Gates Foundation) and Engineering World Health: Projects That Matter. \\ With the advent of wearables and mobile technology, fitness trackers such as Fitbit and Jawbone UP have become increasingly popular. They have been able to provide users with a more detailed look at their daily lives through measurements taken from a small number of basic sensors and well-designed smartphone apps. Despite these successes in the area of personal fitness, the wearable market has not yet successfully expanded into other areas, and only uses very limited forms of machine learning. Through this project, students will look to expand the possibilities of wearables into other areas beyond fitness by working with sensors beyond just the accelerometers and gyroscopes typically found in wearables on the market, as well as look at more sophisticated ways to interpret sensor data and provide users with better insight into their daily lives. |
|
|
NSF EEC:
NSF EEC 1129525: Collaborative Research: Enabling Instructors to Teach Statics Actively PI: Tracy Hammond, CoPIs: Julie Linsey, Erin McTigue 2011-09-01–2017-08-31, $540,970 Abstract: This engineering education research project builds from previous research on how computer-aided feedback to students improves how the students learn drawing-related concepts in engineering statics. By researching how to integrate rapid feedback between instructor and student, this project will enable computer-assisted drawing tools to be utilized in introductory engineering statics courses to improve student learning. The broader significance and importance of this project will be to overcome barriers for integrating effective software tools into critical, large enrollment engineering courses early in the curriculum. The collaborative project engages two universities with different educational missions to ensure that the results are widely applicable. The needs of students typically under-represented in engineering are addressed in the proposed research project. This project overlaps with NSF's strategic goals of transforming the frontiers by enhancing research infrastructure and data access to broaden research capabilities. Additionally NSF's goal of innovating for society is enabled by supporting the development of innovative learning systems. |
|
Texas Department of Aging and Disability Services (TDADS):
Texercise Select: Building the Evidence-Base PI: Marcia Ory 2015-09-01–2017-05-31, $204,172 Abstract: Examine the impact of Texercise Select on increase physical activity and healthy eating, enhanced social interaction, perceived quality of life, and physical functioning. |
|
Future Position X:
Automatic Identification of Health Related Activities such as Drinking, Smoking, and Eating PIs: Daniel Goldberg and Tracy Hammond 2016-05-01–2017-04-30, $25,000 Abstract: Using novel artificial intelligence algorithms developed by Hammond to automatically recognize and predict when participants are smoking, drinking, or eating using only a single accelerometer on the wrist and contextual information from a carried cell phone |
|
|
Future Position X:
Building the Framework for a Spatial Health Marketplace PIs: Daniel Goldberg and Tracy Hammond 2016-05-01–2017-04-30, $74,250 Abstract: In order to facilitate the storage of and research on personalized health, location, and activity information for the population of Sweden, we are building a multi-tiered database, website, and hand-held framework and solution. |
|
|
NSF IIP:
NSF IIP 1546906 I-Corps: Evaluating the Market Potential of the Mechanix Sketch Recognition Software PI: Tracy Hammond 2015-06-01–2016-12-31, $50,000 Abstract: Introductory engineering courses within large universities often have annual enrollments exceeding several hundreds of students, while MOOCS and online classes have even larger classes. It is very challenging to achieve differentiated instruction in classrooms with class sizes and student diversity of such great magnitude. In such classes, professors can only assess whether students have mastered a concept by using multiple-choice questions. However, in a multiple choice scenario, students only have to identify the answer rather than create the answer, and the feedback received is only of a binary nature (right or wrong). Additionally, a growing concern among engineering educators is that students are losing both the critical skill of sketched diagrams and the ability to take a real system and reduce it to an accurate but simplified free-body diagram (FBD). The proposed software program, Mechanix, provides an artificially intelligent, scalable online instructional platform for engineering and physics instruction. Mechanix is a sketch-based tutoring system for engineering students enrolled in statics courses. Mechanix not only allows students to hand-draw solutions with planar truss and free body diagrams, just as they would with pencil and paper, but it also checks the student's work. It uses sketch recognition to determine both the component shapes and features of the sketched diagram and the relationships between those shapes and features. Mechanix then uses those relationships to determine whether a student's work is correct and why it is incorrect, enabling Mechanix to return immediate personalized feedback to the student otherwise not possible in large classes. The iterative correction process facilitates student learning. Additionally Mechanix relieves the instructor of the resource intensive burden of grading even very complex, hand-sketched work. This enables the instructor to explore creative problem solving material that may be otherwise too resource intensive to evaluate. The Mechanix software has been successfully deployed in the classroom at three schools with over 300 participating students, showing promising learning and student engagement results. In this project the I-Corps team seeks to determine customer needs in those markets where Mechanix may have the greatest impact in its current form: learning institutions, instructors, students and providers of educational content. The team also seeks to determine the alignment of Mechanix with those needs and changes that may be required; and to determine whether there exist additional markets where this technology may be applied. |
|
|
Texas A&M Engineering Experiment Station:
Center for Healthy Active Living PI: Tracy Hammond 2014-09-01–2016-08-31, $122,000 Abstract: Rapid advances in sensor technologies, big data analytics and processing, electronic health and medical records, and the realization of the so called “Internet of Things” have created a world in which devices and data can be harnessed to promote health and wellbeing in ways unimaginable just a mere decade ago. The proposed Texas A&M University Center for Healthy Active Living (TAMU-CHAL) would serve as a research, education, outreach, and clinical nucleus within which multi-disciplinary research teams can be empowered to develop, test, and deploy new techniques and strategies which exploit this new ecosystem of data, services, and sensors to improve public and individual health. The core mission of this center would be to engineer technologies and techniques for data collection and analyses, facilitate new forms of health research and analyses, and enable novel lines of health interventions, monitoring, and promotion for targeted populations of at-risk individuals. The proposed center leadership team spans the core disciplines of Computer Science and Engineering, Geographic Information Science, and the Health Sciences, and brings the combined expertise necessary to make this vision a reality. The center team has a track record of innovative collaboration that has already resulted in pilot studies, numerous high-impact papers, and proposals for funding with resources secured. In partnership with Baylor Scott & White Health (a leading healthcare services provider), the TAMU-CHAL would serve to integrate and coordinate the research efforts across the TAMU Colleges of Engineering, Geosciences, Education & Human Development, Medicine, and the School of Public Health, serving as a concrete example under the banner of the TAMU One Health Initiative. Funds are requested to support the development of NIH planning and center grant proposals under the U, P, and R mechanisms. |
|
Microsoft Surface Hub:
Sketch and Gesture Recognition for Collaborative and Design Interfaces on the Surface Hub PI: Tracy Hammond 2015-07-01–2016-06-30, $45,000 Abstract: The surface hub is an 82 inch monitor that supports pen and multi-touch. This project is to use novel algorithms developed by Hammond to support novel educational uses for the SurfaceHub. |
|
DyKnow:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $1,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
Environmental Systems Research Institute (ESRI):
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $1,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
Fujitsu:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $2,500 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
Grahl Software:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $1,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
MIcrosoft OneNote:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $1,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
MIcrosoft Research:
Microsoft Perceptive Pixel PI: Tracy Hammond 2014-01-01–2014-12-31, $7,296 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
Texas A&M University Dean of Faculties:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $5,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
Texas A&M College of Engineering:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $10,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
Texas A&M Engineering Experiment Station:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $10,000 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
WACOM:
Workshop on the Impact of Pen and Touch Technology on Education PI: Tracy Hammond 2013-05-31–2014-12-31, $2,500 Abstract: Pen, touch, and tablets are changing the face of education. Be part of the revolution! The Workshop on the Impact of Pen and Touch Technology in Education (WIPTTE) attempts to bring together users, developers, decision makers, students, and teachers to help ascertain and drive the trend of future education. WIPTTE has been going strong for eight years now, with this being the eight annual instantiation of the conference. Each year approximately 150 participants from industry, academia (both developers and users), K-12 teachers, and junior/high school students came to share their tools, experiences, ideas, innovative uses with this new hands-on technology. |
|
|
NSF DUE:
NSF DUE/CCLI 0942400, Sketched-Truss Recognition Tutoring System PI: Julie Linsey, CoPIs: Tracy Hammond and Erin McTigue 2010-03-01–2013-02-28, $199,769 Abstract: The project is designed to evaluate and refine a sketch recognition tutoring system for trusses that enhances engineering learning by providing intelligent and immediate feedback. Through a study of student learning and the revision of the software based on the study's findings, this work provides a springboard for future development of tutoring systems for a much wider array of engineering courses including, but not limited to thermal sciences and a range of other mechanics-based classes. The software actively engages the student while providing guidance and develops a cognitive scaffold between concepts students understand to deeper levels of learning. Most professors are aware of the importance of open-ended problems for deepening learning and enhancing innovation skills. Intellectual Merit: The STRAT (Sketched-Truss Recognition and Analysis Tool) facilitates the incorporation of open-ended design problems into large traditional classes. STRAT software is designed to provide a more efficient means to teach engineers basic mechanics concepts thus allowing study time to be more productive and additional material to be added to the curriculum. Users draw diagrams as they would naturally and thus there is no steep learning curve associated with this tool unlike tool palette or other CAD-based programs. Furthermore, very minimal hardware is required; a standard computer and mouse is adequate. The project has two goals. The first goal is to measure the effects on student learning of the STRAT software through both quantitative and qualitative methods. The second goal of the project is to refine and improve the STRAT software based on experimental data, customer needs collected from students and professors, and on user feedback. To increase external validity and utility, STRAT is being tested in a classroom setting. Broader Impact: The STRAT software has the potential to profoundly increase student learning and the efficiency of the instructional process. The STRAT tool can be used at a variety of institutions of higher learning ranging from small classrooms at teaching focused schools to public research focused universities with very large classrooms. Visual aids are common in civil and mechanical engineering but active sketching and feedback provides additional benefits. Providing immediate feedback improves learning by eliminating the possibility of reinforcing inaccurate assumptions produced by the learner. The sketch tool also encourages users to actively process the information instead of passively viewing visual representations. The task of free-sketching encourages and demands, that learners actively construct their knowledge leading to higher levels of comprehension and longer term learning. This approach to learning is also particularly suited for the sciences as it encourages the creation of new knowledge rather than the memorization of previously established information. Furthermore, there are plans for the STRAT software to be freely available on a website and results broadly disseminated through conferences and journals. |
|
|
NSF EEC:
NSF EEC 0935219, Civil Engineering Sketch Workbook PI: Tracy Hammond, CoPIs: Anthony Cahill 2009-10-01–2012-09-30, $400,000 Abstract: This engineering education research award to Texas Engineering Experiment Station will employ researchers to develop a novel sketch recognition-based learning system which will allow students to use freehand sketching to enter their drawings of forces acting on components of physical systems. This work will advance understanding of automatic real-time interpretation of combined text and drawings. As a result, engineering learning will be improved by providing students with a natural unconstrained input mechanism and immediate feedback which will reduce the probability of developing misconceptions. In addition, engineering faculty will have new capability to track common student errors. The tool will be deployed in a first-year core engineering course in statics but has application in a wide range of STEM courses. The improved learning in core courses will contribute to better preparation of engineering students for more advanced study. |
|
Office of Naval Research:
SBIR, Multimodal Interface for Unmanned Aircraft Systems Polarity Labs (Lead) PI: Stephane Fymat, TAMU (sub) PI: Tracy Hammond 2010-06-01–2012-05-31, $250,000 Abstract: We propose to develop a multi-modal interaction framework an unmanned helicopter and warfighers who are not skilled UAS operators. This framework can support multiple modes, such as sketch, speech, chat and gesture, including simultaneous multi-modal input in one interaction. Benefits: We anticipate delivering a toolkit that can be used by any manufacturer to enable humans to interact with their unmanned system via our scalable interaction approach, either directly from human to UAS or via a ground control station. This toolkit will include the framework, an API set and a fully implemented set of features for one or two modes, likely to be sketch and speech. |
|
Google Research:
Sketch Recognition: Algorithms, Interfaces, and a Platform for Engineering Education and Beyond PI: Tracy Hammond 2010-01-01–2011-12-31, $50,000 Abstract: The goal of this project is to advance the field of sketch recognition through the development of new algorithms in artificial intelligence that improve recognition accuracy, decrease recognition time, increase the number and types of shapes recognized, and allow for more freeform drawing, including intermixed command gestures, shapes and text. |
|
DARPA, IPTO Office:
CSSG Study: Operation Geotroopers: Gaining Time and Maneuverability on the Drop Zone PI: Tracy Hammond 2010-09-01–2011-08-31, $147,733 Abstract: |
|
|
NSF CRA:
REU support for the Mechanix project PI: Tracy Hammond 2010-09-01–2011-08-31, $25,500 Abstract: |
|
|
NSF CreativeIT CHS:
Creative IT 0757557, Pilot: Let Your Notes Come Alive: The SkRUI Classroom Sketchbook PI: Tracy Hammond, CoPI: Donald Maxwell 2008-06-01–2011-05-31, $247,000 Abstract: This project develops a creativity enhancing tool for innovative education by building an electronic laboratory notebook application for a tablet PC in which students can combine graphical and handwritten notes in electronic form. The graphical diagrams are understood by sketch recognition systems built by the students and their teachers. Students take their electronic laboratory notebooks from class to class; the application studies the user's context to determine the most likely domain in which the student is drawing. This new model of active visualization will increase students' understanding of the graphical material. This tool allows students to draw free-hand drawings, just as they would on paper into a tablet PC, that are automatically recognized by the computer using sketch recognition system that provides simulation, feedback, and search capabilities. The tool will be evaluated by observation and interview to determine if the students are more creative when using the sketching tool. This work will aid the learning process through a new visualization model for real-time simulation of the students' hand-drawn sketches, ultimately being used to teach students at all levels from kindergarten to graduate students. |
|
|
DARPA, Computer Science Study Panel:
Fundamental Risky Research on Sketch Recognition PI: Tracy Hammond 2010-04-10–2011-03-31, $100,000 Abstract: |
|
Rockwell Collins Charitable Contributions University Allocations:
Hand-Tracking Recognition Course PI: Tracy Hammond 2009-07-01–2010-06-30, $30,000 Abstract: The proposal requests materials for a design course on: Haptics Course Description: Analysis, implementation, development, and comparison of haptic devices used for natural input and simulation devices. Interaction devices used in the class include hand-tracking gloves, eye-tracking glasses, pens, and other 3D inputs sensors and devices. Algorithms will look at motion, geometry, speed, as well as other contextual features for recognition. Course material includes an examination of models of classification using different AI techniques including linear classification, manifold learning, graphical models, HMMs, and others. Graduate MS students and senior undergraduate students are invited to participate. |
|
|
DARPA (BAE/SIFT):
Deep Green: Commander’s Associate (BAE/SIFT Team) PI (under BAE/SIFT): Tracy Hammond 2008-04-22–2009-05-31, $461,916 Abstract: Deep Green will develop technologies to help the commander create courses of action (options), fill in details for the commander, evaluate the options, develop alternatives, and evaluate the impact of decisions on other parts of the plan. The permutations of these option sketches for all sides and forces are assembled and passed to a new kind of combat model which generates many qualitatively different possible futures. These possible futures are organized into a graph-like structure. The commander can explore the space of possible futures, conducting “what-if” drills and generating branch and sequel options. Deep Green will take information from the ongoing, current operation to estimate the likelihood that the various possible futures may occur. Using this information, Deep Green will prune futures that are becoming very improbable and ask the commander to generate options for futures that are becoming more likely. In this way, Deep Green will ensure that the commander rarely reaches a point in the operation at which he has no options. This will keep the enemy firmly inside our decision cycle. To accomplish this, Deep Green will focus on 4 functional components: (1) The Commander’s Associate, incl. (a) Sketch to Plan and (b) Sketch to Decide. Automatically converts the commander’s hand-drawn sketch with accompanying speech of his intent into a Course of Action (COA) at the brigade level. Sketch to decide takes feedback from other elements and allows the commander to explore some “what ifs” and probability branches associated with them. (2) Blitzkrieg. Takes the commander’s decisions and generates potential battle outcomes. Note the use of the plural term. Besides being very fast (the blitz in Blitzkrieg), it is intended to generate a broad set of possible futures. These futures should be feasible, even if not expected by human users. (3) Crystal Ball. Receives options from Sketch to Plan. Controls the operation of Blitzkrieg in generating futures. Takes information from the ongoing operation and updates the likelihood metrics associated with possible futures; also generates two additional metrics associated with the possible futures: value/utility and flexibility. Uses those updated likelihood metrics to prune parts of the futures graph and nominate futures at which the commander should generate additional options and invokes Sketch to Plan. Finally, it identifies upcoming decision points and invokes Sketch to Decide. |
|
|
DARPA, SAIC:
Deep Green: Commander’s Associate (SAIC Team) PI (sub under SAIC): Tracy Hammond 2008-04-22–2009-05-31, $461,916 Abstract: Deep Green will develop technologies to help the commander create courses of action (options), fill in details for the commander, evaluate the options, develop alternatives, and evaluate the impact of decisions on other parts of the plan. The permutations of these option sketches for all sides and forces are assembled and passed to a new kind of combat model which generates many qualitatively different possible futures. These possible futures are organized into a graph-like structure. The commander can explore the space of possible futures, conducting “what-if” drills and generating branch and sequel options. Deep Green will take information from the ongoing, current operation to estimate the likelihood that the various possible futures may occur. Using this information, Deep Green will prune futures that are becoming very improbable and ask the commander to generate options for futures that are becoming more likely. In this way, Deep Green will ensure that the commander rarely reaches a point in the operation at which he has no options. This will keep the enemy firmly inside our decision cycle. To accomplish this, Deep Green will focus on 4 functional components: (1) The Commander’s Associate, incl. (a) Sketch to Plan and (b) Sketch to Decide. Automatically converts the commander’s hand-drawn sketch with accompanying speech of his intent into a Course of Action (COA) at the brigade level. Sketch to decide takes feedback from other elements and allows the commander to explore some “what ifs” and probability branches associated with them. (2) Blitzkrieg. Takes the commander’s decisions and generates potential battle outcomes. Note the use of the plural term. Besides being very fast (the blitz in Blitzkrieg), it is intended to generate a broad set of possible futures. These futures should be feasible, even if not expected by human users. (3) Crystal Ball. Receives options from Sketch to Plan. Controls the operation of Blitzkrieg in generating futures. Takes information from the ongoing operation and updates the likelihood metrics associated with possible futures; also generates two additional metrics associated with the possible futures: value/utility and flexibility. Uses those updated likelihood metrics to prune parts of the futures graph and nominate futures at which the commander should generate additional options and invokes Sketch to Plan. Finally, it identifies upcoming decision points and invokes Sketch to Decide. |
|
|
NSF CHS:
NSF IIS 0744150: Developing Perception-based Geometric Primitive-shape and Constraint Recognizers to Empower Instructors to Build Sketch Systems in the Classroom PI: Tracy Hammond 2007-09-01–2008-08-31, $149,858 Abstract: This research attempts to identify and measure the effect of perceptual and contextual features on sketch recognition, and use these results to create effective classifiers to recognize low level shapes and constraints that will identify all possible interpretations along with a ranking for use by higher-level recognition systems. Contextual features to be examined include whether users are drawing or viewing a shape, whether users are viewing the beautified or hand-drawn shape, accompanying hand movements, domain knowledge, and accompanying shapes in the diagram. User studies in perception will determine how geometric features co-vary and how shapes should be varied to agree with human perception. Graphical diagrams are an important part of the educational process. Unfortunately, they are time-consuming to correct and are usually omitted from the testing process despite evidence that testing aids in learning of subject material. Sketch recognition systems can be built to recognize hand-drawn diagrams, but they currently take a long time to build and require expertise in sketch recognition. This project has the potential to provide foundational work that could lead to the development of a tool to allow instructors, without sketch recognition expertise, to build their own sketch recognition tools. Further, this project proposes to build geometric primitive and constraint recognizers based on perception and context to make the creation of sketch recognition systems more intuitive for non-experts in sketch recognition by better matching computer-based recognition to perceptual and contextual expectations. The results from this project will be implemented in the LADDER/GUILD technologies to 1) improve recognition results, making the sketch recognition systems more useful for instructors, and 2) improve automatic generation of shape descriptions to simplify sketch system creation, making it more practical for instructors to use the system. |