Image Classification with Knet

In this article we'll use Julia and the Knet package to classify an image:

Qat.jpg

The Knet project is a machine learning library implemented in Julia at Koç University. With Knet, models are defined using plain Julia code, including many high-level language features which other libraries prohibit. The library uses memory management and efficient GPU kernels to achieve good performance.

We'll load up a Knet implementation of the ResNet Deep Residual Network model, and weights from pretraining for the ImageNet ILSVRC challenge. The network can then be used to classify uploaded images.

1.
Setup

Much of what we need is provided by Nextjournal and its default Julia environment. We'll just have to install Knet and a few dependencies. We currently install a branch of MAT.jl which works on Julia 1.0 but is unoptimized—this should be updated once the relevant PR is merged.

pkg"update"
pkg"add OffsetArrays FFTViews ArgParse Images Knet"
pkg"add https://github.com/halleysfifthinc/MAT.jl#v0.7-update"
pkg"build"
pkg"precompile"

We'll also download the pretraining data to /results.

wget --progress=dot:giga -O /results/imagenet-resnet-152-dag.mat \
    http://www.vlfeat.org/matconvnet/models/imagenet-resnet-152-dag.mat
imagenet-resnet-152-dag.mat

2.
Model Definition

We'll pull from the Knet example files: the ResNet-152 model implemented in resnet.jl. This residual network is similar to a standard layered network, but separate building blocks of a few layers are directly connected via shortcuts. This mixing solves the problem of degradation of training accuracy seen in some deep networks without adding any significant computational complexity.

Figure: Individual building block designs for ResNet models, showing the shortcut from one side of the block to the other. The right example has a 'bottleneck' design used by the ResNet-152 model to reduce training time.

import ArgParse,Images,MAT,ImageMagick,Knet

# mode, 0=>train, 1=>test
function resnet152(w,x,ms; mode=1)
    # layer 1
    conv1 = reslayerx1(w[1:3],x,ms; padding=3, stride=2, mode=mode)
    pool1 = pool(conv1; window=3, stride=2)

    # layer 2,3,4,5
    r2 = reslayerx5(w[4:33], pool1, ms; strides=[1,1,1,1], mode=mode)
    r3 = reslayerx5(w[34:108], r2, ms; mode=mode)
    r4 = reslayerx5(w[109:435], r3, ms; mode=mode)
    r5 = reslayerx5(w[436:465], r4, ms; mode=mode)

    # fully connected layer
    pool5  = pool(r5; stride=1, window=7, mode=2)
    fc1000 = w[466] * mat(pool5) .+ w[467]
end

Then the layer definition functions and some utility functions, including an image processing function from vgg.jl.

0.5s
# Batch Normalization Layer
# works both for convolutional and fully connected layers
# mode, 0=>train, 1=>test
function batchnorm(w, x, ms; mode=1, epsilon=1e-5)
    mu, sigma = nothing, nothing
    if mode == 0
        d = ndims(x) == 4 ? (1,2,4) : (2,)
        s = prod(size(x,d...))
        mu = sum(x,d) / s
        x0 = x .- mu
        x1 = x0 .* x0
        sigma = sqrt(epsilon + (sum(x1, d)) / s)
    elseif mode == 1
        mu = popfirst!(ms)
        sigma = popfirst!(ms)
    end

    # we need getval in backpropagation
    push!(ms, AutoGrad.value(mu), AutoGrad.value(sigma))
    xhat = (x.-mu) ./ sigma
    return w[1] .* xhat .+ w[2]
end

function reslayerx0(w,x,ms; padding=0, stride=1, mode=1)
    b  = conv4(w[1],x; padding=padding, stride=stride)
    bx = batchnorm(w[2:3],b,ms; mode=mode)
end

function reslayerx1(w,x,ms; padding=0, stride=1, mode=1)
    relu.(reslayerx0(w,x,ms; padding=padding, stride=stride, mode=mode))
end

function reslayerx2(w,x,ms; pads=[0,1,0], strides=[1,1,1], mode=1)
    ba = reslayerx1(w[1:3],x,ms; padding=pads[1], stride=strides[1], mode=mode)
    bb = reslayerx1(w[4:6],ba,ms; padding=pads[2], stride=strides[2], mode=mode)
    bc = reslayerx0(w[7:9],bb,ms; padding=pads[3], stride=strides[3], mode=mode)
end

function reslayerx3(w,x,ms; pads=[0,0,1,0], strides=[2,2,1,1], mode=1) # 12
    a = reslayerx0(w[1:3],x,ms; stride=strides[1], padding=pads[1], mode=mode)
    b = reslayerx2(w[4:12],x,ms; strides=strides[2:4], pads=pads[2:4], mode=mode)
    relu.(a .+ b)
end

function reslayerx4(w,x,ms; pads=[0,1,0], strides=[1,1,1], mode=1)
    relu.(x .+ reslayerx2(w,x,ms; pads=pads, strides=strides, mode=mode))
end

function reslayerx5(w,x,ms; strides=[2,2,1,1], mode=1)
    x = reslayerx3(w[1:12],x,ms; strides=strides, mode=mode)
    for k = 13:9:length(w)
        x = reslayerx4(w[k:k+8],x,ms; mode=mode)
    end
    return x
end
0.4s
function get_params(params, atype)
    len = length(params["value"])
    ws, ms = [], []
    for k = 1:len
        name = params["name"][k]
        value = convert(Array{Float32}, params["value"][k])

        if endswith(name, "moments")
            push!(ms, reshape(value[:,1], (1,1,size(value,1),1)))
            push!(ms, reshape(value[:,2], (1,1,size(value,1),1)))
        elseif startswith(name, "bn")
            push!(ws, reshape(value, (1,1,length(value),1)))
        elseif startswith(name, "fc") && endswith(name, "filter")
            push!(ws, transpose(reshape(value,(size(value,3),size(value,4)))))
        elseif startswith(name, "conv") && endswith(name, "bias")
            push!(ws, reshape(value, (1,1,length(value),1)))
        else
            push!(ws, value)
        end
    end
    map(wi->convert(atype, wi), ws),
    map(mi->convert(atype, mi), ms)
end

# From vgg.jl
function data(img, averageImage)
    if occursin("://",img)
        @info "Downloading $img"
        img = download(img)
    end
    a0 = load(img)
    new_size = ntuple(i->div(size(a0,i)*224,minimum(size(a0))),2)
    a1 = Images.imresize(a0, new_size)
    i1 = div(size(a1,1)-224,2)
    j1 = div(size(a1,2)-224,2)
    b1 = a1[i1+1:i1+224,j1+1:j1+224]
    c1 = permutedims(channelview(b1), (3,2,1))
    d1 = convert(Array{Float32}, c1)
    e1 = reshape(d1[:,:,1:3], (224,224,3,1))
    f1 = (255 * e1 .- averageImage)
    g1 = permutedims(f1, [2,1,3,4])
end

3.
Classification

Let's see how this model does. As is tradition, we shall cat.

We'll wrap the main model-handling part of the code in a predict() function, so that the frontend won't be trying to display a 240 MB array—otherwise, this is fairly straightforward.

40.4s
using Knet, MAT, Images

o = Dict(
  :atype => KnetArray{Float32},
  :model => imagenet-resnet-152-dag.matdata download 152,
  :image => Image,
  :top   => 10
)

function predict(o)
	@info "Reading $(o[:model])"
	model = matread(abspath(o[:model]))
	avgimg = model["meta"]["normalization"]["averageImage"]
	avgimg = convert(Array{Float32}, avgimg)
	description = model["meta"]["classes"]["description"]
	w, ms = get_params(model["params"], o[:atype])

	@info "Reading $(o[:image])"
	img = data(o[:image], avgimg)
	img = convert(o[:atype], img)

	@info "Classifying."
	@time y1 = resnet152(w,img,ms)
  
  return y1, description
end

y1, description = predict(o)

z1 = vec(Array(y1))
s1 = sortperm(z1,rev=true)
p1 = exp.(logp(z1))
using Printf

for ind in s1[1:o[:top]]
  print("$(description[ind]): $(@sprintf("%.2f",p1[ind]*100))%\n")
end

Well, it's pretty sure there's some type of cat! It also has an inkling that there's some sort of textile in there. Nifty.

© 2018 Nextjournal GmbH