Jason Bunk's repositories
tensorflow-bilateral-permutohedral
Port of permutohedral bilateral filtering to tensorflow as an op
densecut-linux
DenseCut ported to Linux (now cross-platform between Windows and Linux) -- from http://mmcheng.net/densecut/
reshade_cv
(WIP) 3D datasets from games - RGBD, NeRFs, point clouds
NBodyGalaxySimulation
N-body simulation of colliding galaxies: final project for UCSD Physics 141 and Physics 244, 2015. Jason Bunk.
BunkNeuralNets
My neural network library for educational purposes - from scratch, using Numpy with Python (CPU only).
BayesianRNN
Code for the paper "A Theoretically Grounded Application of Dropout in Recurrent Neural Networks"
Bonsai
Bonsai GPU tree code
caffe
Latest Caffe merged with CRF-RNN (GPU), py-faster-rcnn, and other additions
ccv
C-based/Cached/Core Computer Vision Library, A Modern Computer Vision Library
char-rnn-tensorflow
RNNs with dropout in the hidden state for approximate Bernoulli variational posterior for weights
cod-cnn
Curse of Dimensionality (CoD) and Convolutional Neural Networks (CNN)
experiment-backer-upper
Backs up the state of your code and config files every time you start a machine learning experiment
InvertedPendulumVisualControl
Computer vision for feedback measurements for an inverted pendulum control system, using SFML for cross-platform visualization and interactivity, and JPhysics for real time modeling.
JPhysics
Object-oriented C++ numerical methods library I started in high school, with broad applications from 2D video game physics to n-body simulations to robotics. Includes visualization with OpenGL.
JSolarSystem
Realistic 3D interactive solar system simulation with n-body physics using JPhysics. Can be used for general numerical n-body simulations of complexity O(N^2) by changing the initial conditions and types of bodies (used for class projects in computational physics at UCSD).
python-script-to-jupyter-notebook
Convert python scripts to Jupyter notebooks, while converting argparse to a notebook-friendly cell
svae
code for Structured Variational Autoencoders
TensorFlow-VAE-GAN-DRAW
A collection of generative methods implemented with TensorFlow (Deep Convolutional Generative Adversarial Networks (DCGAN), Variational Autoencoder (VAE) and DRAW: A Recurrent Neural Network For Image Generation).