Portfolio

Portfolio

Below you can find a selection of projects particularly relevant to my current interests, including ones that I have worked on in my free-time and in larger teams. As always, you can find more on my LinkedIn and GitHub profile.


Industry Experience

Vision-Based Tracking System for Ground-Truthing, iRobot.

Acted as primary developer and designer for upgrading the robot tracking system used to test and robot behavior and provide ground truth for on-robot localization algorithms. Developed Unscented Kalman filter in Python, C++, and ROS to handle non-linear estimation and sensor fusion.

Research Experience

Robotic Finger, Duke Intelligent Motion Laboratory.

Developed a 3-joint microcontroller-based robotic manipulator in C, with easy to use Python API. Created a PC–Microcontroller data communication library to provide easy communication between microcontroller and Python (without having to worry about buffer overflow). Manipulator was integrated with the “Tele-Robotic Intelligent Nursing Assistant“,  a robot designed to let nurses remotely perform common tasks while caring for patients under quarantine.

Personal Experience

Highway Lane Tracker.

Tracks the position of a car within lane lines using computer vision. Robust to gaussian and non-gaussian noise as well as signal dropout. Find a more in-depth discussion at my GitHub and the blog write-up.

Behavioral Cloning Car.

Created a neural network that autonomously steers a simulated vehicle around limited-visibility tracks using camera sensors only. Clones the behavior of human drivers after training on human driving dataGithub.

Vehicle Detector.

Detects (and places bounding boxes around) cars in video using an SVM and sliding window search. Github.

System Integration for Animal Tracking Drone.

Led system integration & development of UAV capable of locating visually tagged animals mid-flight. Took advantage of multiple open source libraries, combining them into one cohesive platform allowing for continued development of remote sensing applications. Project Website.

Nvidia Jetson CSI Camera Launcher for ROS.

A ROS package making it simple to use CSI cameras on the Nvidia Jetson TK1, TX1, or TX2 with ROS. Github.

Road Sign Classifier.

Classifies German traffic signs using convolutional neural networks. Github.