Intro / Motivation / Magnify 360 screenshots (proably not the Facebook Screenshots down to here

1. 2008-2009 Magnify360 Computer Science ClinicOptimizing Individual Web Experiences Magnify 360 logo Treating visitors as individuals User Experience Evaluation Underlying System These robots were shown with Mapping for All, an exhibit at AAAI 2008. At left is the One Robot Per Child project, where an OLPC provides control and sensing. Intro / Motivation / Magnify 360 screenshots (proably not the Facebook Screenshots down to here Because of the markers' format, color dominates the visual processing in our system. The vision software allows for quick recalibration of the color definitions that distinguish one object from another. Once defined, the colors provide distinct patches from which connected components are extracted. Vertically aligned components form markers. One of the sensitivity graphs the table Caption Caption for the three screenshots and labels, above These robots were shown with Mapping for All, an exhibit at AAAI 2008. At left is the One Robot Per Child project, where an OLPC provides control and sensing. caption Explanation robot combines several partial behaviors, e.g., for wall-following, marker detection and approach, and recovery from unexpected obstacle bumps. A finite state machine arbitrates among these behaviors based on the current sensory input and current goals. Feedback is available to observers through the laptop's voice synthesizer and/or network access from a second machine. • Explain stuff • bullet list 1 • bullet list 2 • bullet list 3 Results From clickstream data: Profiles caption An unusual feature of our robot, Import Antigravity, is that its computation comes from an ordinary Mac laptop held by a custom-designed platform. The advantage of using off-the-shelf computation is that networking, the development environment, and interface with ordinary peripherals requires no additional setup. For example, the Create itself is simply a serial peripheral of the Mac it is carrying. Clustering Users Classification and Testing caption caption Explanation robot combines several partial behaviors, e.g., for wall-following, marker detection and approach, and recovery from unexpected obstacle bumps. A finite state machine arbitrates among these behaviors based on the current sensory input and The range sensors available on the simulated robot were lasers with 1 degree of angular resolution. with that much data, it is possible to recognize locations, for example, through the range signatures of each of the four corners of the room. Caption Explanation robot combines several partial behaviors, e.g., for wall-following, marker detection and approach, and recovery from unexpected obstacle bumps. A finite state machine arbitrates among these behaviors based on the current sensory input and These robots were shown with Mapping for All, an exhibit at AAAI 2008. At left is the One Robot Per Child project, where an OLPC provides control and sensing. and here caption Acknowledgments Caption for the table and profile graphics and labels, above Magnify360 Liaisons David Lapayowker '09 Marissa Quitt '09 Elaine Shaver '09 (PM) Devin Smith '09 Team Members David Lapayowker '09 Marissa Quitt '09 Elaine Shaver '09 (PM) Devin Smith '09 Explanation robot combines several partial behaviors, e.g., for wall-following, marker detection and approach, and recovery from unexpected obstacle bumps. A finite state machine arbitrates among these behaviors based on the current sensory input and An unusual feature of our robot, Import Antigravity, is that its computation comes from an ordinary Mac laptop held by a custom-designed platform. The advantage of using Faculty Advisor Zachary Dodds Caption