Texas A&M University Sketch Recognition Lab Computer Science Department  

Tracy Anne Hammond

Office: HRBB 414C
Email: hammond at cse.tamu.edu
Phone: 979/862-4284 Fax: 979/847-8578

Associate Professor and Director of the

Sketch Recognition Lab

Courses Research Publications


Do you need more course credits? Are you an undergraduate or a graduate student interested in working on a real research project over this summer for course credit? Join the Summer of Civil Sketch!

Research Interests:

sketch recognition, perception, cognitive behavior, computer human interaction, artificial intelligence, concept learning, computer graphics, psychology, anthropology, the gender gap in computer science


Ph.D. (Computer Science), Massachusetts Institute Technology
Finance Technology Option (FTO), Massachusetts Institute Technology
M.A (Anthropology), Columbia University
M.S. (Computer Science), Columbia University
B.A. (Mathematics), Columbia University
B.S. (Applied Mathematics), Columbia University

Overarching Research Question

Can a computer recognize a human's intention if we allow that human to interact naturally with the world, where ''naturally'' implies effectively as well as harnessing skills already learned through human-human interaction without losing effectiveness? Further, can a computer use that knowledge of intention to empower the human?

To this end, my enacted research thus far has focused and made significant progress on three sub-problems:
  1. Sketch recognition: Can computers understand freely-drawn hand-drawn sketches drawn by a human, as well as, or possibly, even better than, another human can? How can this understanding be used to empower the human?
  2. Gesture recognition: Can computers understand hand gestures made by a human to empower the human?
  3. New Devices/Modes of Human-Computer Interaction: What new modes or devices of interaction could exist to better human-computer interaction?

Research on Sketch Recognition

Sketching is a natural way of devising and communicating ideas. Graphical diagrams can convey ideas that would be difficult to describe using only text. Graphical diagrams pervade education, being an important, if not vital part of multidisciplinary fields such as economics, mechanical engineering, electrical engineering, document editing, and even Asian languages. Pen and paper provide a freedom of interaction that is still preferred to a computer automated design tool, even though users want the sophistication of analysis and simulation that a computer-understood diagram can provide.

Sketch recognition is the automated understanding of a drawn diagram, attempting to recognize the intent of the user while allowing the user to draw in an unconstrained manner. Sketch recognition combines advances in artificial intelligence with the natural interaction of a pen and paper. The automated understanding of naturally drawn diagrams has clear benefits across education, design, and human-computer interaction.

2004 video showing general sketch recognition results developed by Hammond applied to the domain of Mechanical Engineering

Research on Gesture Recognition

Gesture recognition is the automated recognition of hand-gestures. My published research in this area focuses on automatic activity recognition in the office using only the hand posture. Current and future research includes using eye-gaze to help recognize, but more importantly segment continuous hand gestures into expressive, commanding, and non-meaningful gestures.

New Devices for / Modes of Human-Computer Interaction

Coming soon...

Joining the Sketch Recognition Lab / Research Position / Assistantships

Assistantship positions in my group are very selective. As part of that selection process all students are required to first have taken a course with me and excelled in that course before such a position is possible. The course functions as a trial period to determine if our personalities fit well together, and is an advantage to both the student and me.

Writing Guidelines

Blog Summary and Discussion Guidelines
Course Paper Guidelines
Thesis Proposal Guidelines