15-494 Cognitive Robotics
Spring 2012
Course Links:  Main Lectures Labs Homeworks

Tekkotsu Links:   Wiki Tutorial Resources Bugs CVS

Final Project Ideas

  • Support for Kinect skeletons (Serene Han): Extend Tekkotsu's Kinect support to include the OpenNI skeleton tracking facility.

  • Mirage support for Kinect (Mark Perez): Add depth imaging to the Mirage simulator so we can better simulate the Kinect.

  • Face tracker (Andrew Sheng): Port the OpenCV face tracker to Tekkotsu and integrate it with the Lookout and MapBuilder.

  • Accessibility (Bruce Hill and James Criscuolo): Develop enhancements to Tekkotsu to make it more usable by persons with disabilities, such as color blindness, low vision, or hearing impairment.

  • Voice input (David Hamilton, Songdet Nillasithanukroh, and Jinxue Zhou): Provide voice input to Tekkotsu using an Android phone and Google's speech-to-text service, so the robot can respond to commands such as "move forward".

  • Verbal instruction (Chase Brownell, Samuel Russell, and Kelsey Lee): Develop a language-based interface to the robot so humans can refer to things the robot perceives, shift its focus of attention, and command it to pick up objects.

  • Object recognition (Elmer Garduno and Apoorv Khandelwal): Either port the SURF algorithm to Tekkotsu, or fix up the existing SIFT code and get it working reliably.

  • Task planning (Andrew Bauer): Implement a combinatorial task, such as duplicating a pattern of objects, using Tekkotsu's interface to the FF task planner.

Additional Ideas for Next Year:

  • Grasper enhancement: Extend the Grasper to support the Calliope5KP's 5-dof arm. Currently the Grasper only supports simpler arms and just a few types of actions. Additional actions might include stacking blocks, or arranging blocks into specific patterns.

  • Calliope face: Construct a "face" for the Calliope robot using microservos to actuate lips, eyebrows, and/or ears, and LEDs to provide additional features.

  • Robot cooperation: Develop a Mirage simulation in which two robots cooperate to perform a task.

Students are also welcome to propose their own ideas for projects, but the must make a convincing argument that the idea is achievable in the time allotted.