Used supervised learning of the MNIST data set to reach ~84% accuracy when identifying hand written digits.
GooMPY is a Python-based interface to the Google Static Maps API. This script automatically downloads and stitches together map tiles into a single image, allowing for dynamic zooming and panning. GooMPY was later integrated into Professor Simon D. Levy’s desktop ground control station.
Co-created an Android app for the configuration and testing of quadcopters. The app monitors the output of key sensors (e.g., IMU and GPS) as well as input from the controller.
Used GPS signals to navigate drone within 10 feet of 3’ x 3’ target. The algorithm then uses filtered camera input, to identify and slowly descend upon predetermined target.
Created a model of a quadcopter in Python using Tkinter. This model was then used in a desktop ground control station to visualize the attitude (pitch, roll, yaw) being read by the IMU. This allows the user to confirm the sensor's readings align with the actual attitude.
Co-created a Python script that would generate Python, Java, or C++ code based on a JSON specification. The generated code would then parse the specified signals.