Kyle A. Pearson

Data-VisualizAR. Visualize data or 3D models (.obj) in augmented reality through the web with Google Poly, AR.js, flask.py (Mobile compatible)

Artificial-Art. Create animated art pieces with machine learning using a DCGAN. Procedurally generate tiles for any type of mosaic.

Neural-Nebula. A deep conv. generative adversarial network trained on nebula and images of space

Signal-Alignment. Algorithms to align 1D signals via cross correlation and likelihood maximization.

Exoplanet-Artificial-Intelligence. Machine learning is used to search for signals of transiting exoplanets in planetary search surveys (e.g. Kepler and TESS).

Nbody-AI. A python 3 package for generating N-body simulations, computing transit timing variations (TTV) and retrieving orbit parameters from TTV measurements within a Bayesian framework using machine learning

pyduino_webapi. Instructable tutorial on controlling an arduino device via a web api in python

Exoplanet-Light-Curve-Analysis. A python package for modeling exoplanet light curves with nested sampling.

brave_ad_clicker. A Python bot to automatically close ads on the Brave browser

pyduino_leapmotion. Instructable tutorial for building a hand tracking robotic arm using python, arduino and a leap motion controller.

Mars-Machina. A deep conv. variational autoencoder is trained on digital terrain maps of Mars from HiRise/MRO. 3D surfaces are procedurally generated from the latent space in Unity

HiRISE-3D-Model. A python tool for creating 3D Models (.obj, .glb, .gltf) of terrain maps in Blender

PhotometryTools. Tools for performing aperture and PSF photometry in real time

Unity-Variational-Autoencoder. A variational autoencoder made with tensorflow loaded into Unity for the procedural generation of 3D surfaces

Planetary-Collision-VR. 3D visualizations of planetary collisions in virtual reality with SteamVR and Unity.

Rayleigh-Scattering. Empirical fits to Rayleigh scattering cross section measurements in historic data of H2, N2 and CO2 (~300-900 nm).

Exo-Machina. A large language model, GPT-2, is trained on ~1.6M manuscript abstracts from the ArXiv. Fine-tuning is performed with the abstracts from NASA's ADS and the model is later used to generate text for a predictive keyboard

Mars-Brain-Coral-Network. A pipeline for quickly vetting and segmenting JPEG2000 images from HiRISE/MRO using a hybrid approach with a classifier in the Fourier domain and a U-net model in the spatial domain.

dilation-game. C

G-Force-Flow. In this physics based platformer you must traverse an artificial gravity field inside a spatially tessellated void. Convert gravitational potential energy into kinetic energy as you swing, jump and phase your way through obstacles.

GeneticNeuralNetwork. A custom genetic algorithm is used to optimize the architecture of a deep neural network

Exoplanet-Mapping. A Spitzer IRAC data reduction pipeline and 3D exoplanet brightness model using spherical harmonics

Exoplanet-Ionization. Compute alkali abundances in exoplanetary atmospheres subject to ionizing radiation.

Animated-LED-Strip. A programmable LED strip with animations from a Gemma M0 microcontroller and CircuitPython

Exoplanet-Transit-Periodogram. A custom transit periodogram which includes limb darkening effects. A transit model is quickly optimized using a linear least sq approach in order to derive a robust S/N measurement for a range of periods, durations and mid-transit epochs.

3D-Abstract-Frames. A simple smart contract and a python script for minting non-fungible tokens (ERC 721) using the Polygon blockchain

Art-with-Depth. a collection of A.I. generated images with depth maps rendered on the web with p5js + glsl shader

Stock-Suite. A collection of APIs and algorithms to trade stocks

Exoplanet-Mass-Fractionation. Atmospheric escape calculation for the crossover mass at which a heavier second constituent will not be dragged along with a lighter constituent (i.e. H).

Mars-Rock-Detector. Python code for measuring the frequency and size distribution of boulders on Mars using data from HiRISE/MRO. We developed an interactive application for making training data quickly for machine learning applications.