This repository creates all the code, neural networks, and data created in the process of my senior capstone project.
Defines the correct node to publish an offset odometry using a pre-trained neural network with sensor values obtained at the time of each collision.
Contains all of the data collected to train the neural network along with each version of the processed data.
Defines the collect node to faciliate the data collection process.
Contains the code used to split, normalize/standardize, and perform PCA on the data.
Contains the neural networks trained in the project (temp stores networks during parameter search).
Contains code to train neural networks and perform parameter searches.
Contains the outputs from feeding testing data to trained neural networks.
Defines drive node to control robot's movement and publish inputs.