more bounds

.github/workflows/CompatHelper.yml

@@ -0,0 +1,16 @@
+name: CompatHelper
+
+on:
+ schedule:
+ - cron: '00 00 * * *'
+
+jobs:
+ CompatHelper:
+ runs-on: ubuntu-latest
+ steps:
+ - name: Pkg.add("CompatHelper")
+ run: julia -e 'using Pkg; Pkg.add("CompatHelper")'
+ - name: CompatHelper.main()
+ env:
+ GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+ run: julia -e 'using CompatHelper; CompatHelper.main()'

.github/workflows/ci.yml

@@ -7,9 +7,11 @@ jobs:
  name: Julia ${{ matrix.version }} - ${{ matrix.os }} - ${{ matrix.arch }}
  runs-on: ${{ matrix.os }}
  strategy:
+ fail-fast: false
  matrix:
  version:
  - '1.3'
+ - '1'
  - 'nightly'
  os:
  - ubuntu-latest
@@ -23,5 +25,16 @@ jobs:
  with:
  version: ${{ matrix.version }}
  arch: ${{ matrix.arch }}
+ - name: Cache artifacts
+ uses: actions/cache@v1
+ env:
+ cache-name: cache-artifacts
+ with:
+ path: ~/.julia/artifacts 
+ key: ${{ runner.os }}-test-${{ env.cache-name }}-${{ hashFiles('**/Project.toml') }}
+ restore-keys: |
+ ${{ runner.os }}-test-${{ env.cache-name }}-
+ ${{ runner.os }}-test-
+ ${{ runner.os }}-
  - uses: julia-actions/julia-buildpkg@latest
  - uses: julia-actions/julia-runtest@latest

Project.toml

@@ -1,21 +1,25 @@
 name = "UNet"
 uuid = "0d73aaa9-994a-4556-95d0-da67cb772a03"
 authors = ["dhairyagandhi <[email protected]>"]
-version = "0.1.0"
+version = "0.2.0"
 
 [deps]
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
+FileIO = "5789e2e9-d7fb-5bc7-8068-2c6fae9b9549"
 Flux = "587475ba-b771-5e3f-ad9e-33799f191a9c"
-Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
+ImageCore = "a09fc81d-aa75-5fe9-8630-4744c3626534"
+ImageTransformations = "02fcd773-0e25-5acc-982a-7f6622650795"
 Reexport = "189a3867-3050-52da-a836-e630ba90ab69"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 
 [compat]
-Distributions = "0.20"
-Flux = "0.10"
-Images = "0.22"
-Reexport = "0.2.2"
+Distributions = "0.20, 0.21, 0.22, 0.23"
+FileIO = "1"
+Flux = "0.10, 0.11"
+ImageCore = "0.8"
+ImageTransformations = "0.8"
+Reexport = "0.2"
 StatsBase = "0.30"
 julia = "1.3"
 

README.md

@@ -30,14 +30,21 @@ UNet:
  UNetUpBlock(256, 64)
 ```
 
-To default input channel dimension is expected to be `1` ie. grayscale. To support different channel images, you can pass the `channels` to `Unet`.
+The default input channel dimension is expected to be `1` ie. grayscale. To support different channel images, you can pass the `channels` to `Unet`.
 
 ```julia
 julia> u = Unet(3) # for RGB images
 ```
 
 The input size can be any power of two sized batch. Something like `(256,256, channels, batch_size)`.
 
+The default output channel dimension is the input channel dimension. So, `1` for a `Unet()` and e.g. `3` for a `Unet(3)`.
+The output channel dimension can be set by supplying a second argument:
+
+```julia
+julia> u = Unet(3, 5) # 3 input channels, 5 output channels.
+```
+
 ## GPU Support
 
 To train the model on UNet, it is as simple as calling `gpu` on the model.

src/UNet.jl

@@ -7,7 +7,9 @@ using Reexport
 using Flux
 using Flux: @functor
 
-using Images
+using ImageCore
+using ImageTransformations: imresize
+using FileIO
 using Distributions: Normal
 
 @reexport using Statistics

src/dataloader.jl

@@ -6,9 +6,9 @@ using StatsBase: sample, shuffle
 function load_img(base::String, stub::String; rsize = (256,256))
  im = joinpath(base, stub * "_im.tif")
  cr = joinpath(base, stub * "_cr.tif")
- x, y = Images.load(im), Images.load(cr)
- x = Images.imresize(x, rsize...)
- y = Images.imresize(y, rsize...)
+ x, y = load(im), load(cr)
+ x = imresize(x, rsize...)
+ y = imresize(y, rsize...)
  x, y = channelview(x), channelview(y)
  x = reshape(x, rsize..., 1, 1)
  y = reshape(y, rsize..., 1, 1)
@@ -55,10 +55,10 @@ function load_batch(base::String, name_template = "im",
  x = zeros(Float32, rsize..., channels, n) # []
  y = zeros(Float32, rsize..., channels, n) # []
  for (i,(img, mask)) in enumerate(batch)
- img = Images.load(joinpath(base, dir, img))
- mask = Images.load(joinpath(base, dir, mask))
- img = Images.imresize(img, rsize...)
- mask = Images.imresize(mask, rsize...)
+ img = load(joinpath(base, dir, img))
+ mask = load(joinpath(base, dir, mask))
+ img = imresize(img, rsize...)
+ mask = imresize(mask, rsize...)
  img = channelview(img)
  mask = channelview(mask)
  img = reshape(img, rsize..., channels)

src/model.jl

@@ -33,9 +33,10 @@ function (u::UNetUpBlock)(x, bridge)
 end
 
 """
- Unet(channels::Int = 1)
+ Unet(channels::Int = 1, labels::Int = channels)
 
- Initializes a [UNet](https://arxiv.org/pdf/1505.04597.pdf) instance with the given number of channels, typically equal to the number of channels in the input images.
+ Initializes a [UNet](https://arxiv.org/pdf/1505.04597.pdf) instance with the given number of `channels`, typically equal to the number of channels in the input images.
+ `labels`, equal to the number of input channels by default, specifies the number of output channels.
 """
 struct Unet
  conv_down_blocks
@@ -45,7 +46,7 @@ end
 
 @functor Unet
 
-function Unet(channels::Int = 1)
+function Unet(channels::Int = 1, labels::Int = channels)
  conv_down_blocks = Chain(ConvDown(64,64),
  ConvDown(128,128),
  ConvDown(256,256),
@@ -64,7 +65,7 @@ function Unet(channels::Int = 1)
  UNetUpBlock(512, 128),
  UNetUpBlock(256, 64,p = 0.0f0),
  Chain(x->leakyrelu.(x,0.2f0),
- Conv((1, 1), 128=>channels;init=_random_normal))) 
+ Conv((1, 1), 128=>labels;init=_random_normal))) 
  Unet(conv_down_blocks, conv_blocks, up_blocks)
 end
 

test/runtests.jl

@@ -9,6 +9,10 @@ using UNet.Flux, UNet.Flux.Zygote
 
  @test size(u(ip)) == size(ip)
  end
+
+ u = Unet(2,5)
+ ip = rand(Float32, 256, 256, 2, 1)
+ @test size(u(ip)) == (256, 256, 5, 1) 
 end
 
 @testset "Variable Sizes" begin