HEART is an open source platform for personalized telehealth at population scale based on data from internet-connected health sensors and data extracted from the electronic medical record.
Health sensors include scales, blood pressure monitors, activity trackers, continuous glucose monitors, and insulin pumps. HEART has five independent modules: 1) a data processing module that
pulls data in from a variety of devices, 2) an algorithm module that identifies reasons why a provider may want to contact a patient and ranks patients for contact, 3) a visual
interface module that summarizes population data and provides additional details for patients selected by the provider, 4) an intervention module that facilitates interventions, and 5) a user tracking module that monitors providers use.
Various approximate (Bayesian) inference techniques for estimating both epistemic and aleatoric uncertainty in
deep neural networks implemented in eras and TensorFlow (to be officially released by end of May).
Online continuous occupancy mapping with epistemic uncertainty (Numpy and
A simple 2D LIDAR simulator for dynamic environments. We can define static/moving objects and specify/draw the robot's path on a GUI. The simulator outputs 2D pointcloud data coming from a LIDAR. (Python)
Learning nonstationarity in Bayesian Hilbert maps (TensorFlow and Edward)
Learning arbitrary RKHS kernels (Python)
Online domain adaptation for occupancy mapping using Optimal Transport
Modeling spatiotemporal epistemic uncertainty with various likelihood
Robot exploration using scalable uncertainty maps (PyTorch)