Skip to content

Instantly share code, notes, and snippets.


Brandon B outlace

View GitHub Profile
outlace / slide.jl
Created Sep 7, 2021
Julia prototype implementation of Sub-linear Deep Learning Engine (SLIDE); missing advanced hash table updating and other optimizations
View slide.jl
using Flux
using Zygote
using MLDatasets
using LSHFunctions
using DataStructures
using Plots;
using Profile;
using StatProfilerHTML;
using LinearAlgebra;
outlace /
Created Apr 12, 2017
Augments data for machine learning. Takes images in numpy format and randomly applies minor transformations to amplify image data.
import numpy as np
from scipy import ndimage
# Images are assumed to be uint8 0-255 valued.
# For augment function:
# images shape: (batch_size, height, width, channels=3)
# labels shape: (batch_size, 3)
def addBlotch(image, max_dims=[0.2,0.2]):