Object Detection and Segmentation Using Deformable Shape ModelsThis Master's Thesis is in development for the fulfillment of a Master's Degree in Computer Science. The purpose of this thesis is to develop and evaluate an improved deformable shape model and to construct a system to perform ob ject segmentation on images using this model. More information can be found here.
Massively Parallel Processing with Peer-to-Peer NetworksA framework to do distributed massively parallel problems on a peer-to-peer network. Key elements include efficient communication, scalability, performance and fault tolerance. The framework uses the Pastry overlay for node communication. More information can be found here.
Application Level MulticastingAn experiment with SplitStream, a cooperative, distributed, application level multicasting protocol. More information can be found here.
FlockA 3-dimensional simulation illustrating "flocking" and steering behaviors.
Biologically Inspired Object RecognitionA system for recognizing objects in images based on the experimental behavior of the V1-V4 layers of the visual cortex.
Obstacle Detection and Avoidance in Autonomous NavigationA survey on methods used in robotics to detect and maneuver in an environment containing obstacles. The paper can be found here.
A Parts-Based Approach to Object DetectionA combination of top-down and bottom-up approaches to object detection in image scenes. The parts-based approach attempts to construct a descriptor of the object based on "interesting" parts. A voting mechanism is used to find these parts in a scene and segment the object.
MastermindAn artificial intelligence engine which solves the Mastermind game in a number of different ways. A zip file containing the code can be found here.
MIPS-Based Pipelined Processor SimulationDesigned and implemented a simulation of a MIPS-based pipelined processor.