Our team was inspired to create Nabuse because of the increasing number of physical and sexual abuse victims. Nabuse is a wearable device that measures EDA and blood pressure and when certain abnormalities are detected, they contact emergency contacts and 911.
Abnormalities have different measures such as frequency of peaks, deviation from norms, and duration between repetitions.
The bracelets and anklets are made of stretchy and environmentally safe material and the rings are made of stainless metal. They have a sensor on the bottom to measure EDA and blood pressure. When any of the parameters experience abnormalities, they start to vibrate and send a signal to the phone. It vibrates for 15 seconds and in this time duration, the user has the option to turn it off if the detection was not caused by external abuse.
The prototype of the wearables was designed in Blender and the app interface was designed in proto.io. We used Python to utilize data soon to be recorded by users so predict their personal 'norms'. The code utilizes SVM with a linear kernel for 5-fold cross-validation on a dataset loaded from a JSON file. It separates features and labels, builds the SVM model, evaluates it using cross-validation, and prints the mean accuracy along with the standard deviation of the cross-validation scores.
The challenges our team ran into were the different abnormal situations that can happen in the scenario of abuse.
Log in or sign up for Devpost to join the conversation.