YOLO.jl

Currently only supports loading YOLOv2-tiny and the VOC-2007 pretrained model (pretrained on Darknet).

Made possible by Yavuz Bakman's YoloV2

Consider also checking out:

• ObjectDetector.jl -> A Flux-based implementation of YOLO (testing only)
• Darknet.jl -> A julia wrapper of AlexeyAB's fork of Darknet (testing only)

See below for examples or ask questions on

Platform	Build Status
Linux & MacOS x86
Windows 32/64-bit
Linux ARM 32/64-bit
FreeBSD x86

Installation

The package can be installed with the Julia package manager. From the Julia REPL, type ] to enter the Pkg REPL mode and run:

pkg> add YOLO

If you have a CUDA-supported graphics card, make sure that you have CUDA set up such that it satisfies CUDAapi.jl or CuArrays.jl builds.

If you just want to run on CPU (or on a GPU-less CI instance) Knet.jl is currently dependent on a system compiler for the GPU-less conv layer, so make sure you have a compiler installed: i.e. apt-get update && apt-get install gcc g++ for linux or install visual studio for windows

Example Usage (WIP)

Testing a dataset

using YOLO

#First time only (downloads 5011 images & labels!)
YOLO.download_dataset("voc2007")

settings = YOLO.pretrained.v2_tiny_voc.load(minibatch_size=1) #run 1 image at a time
model = YOLO.v2_tiny.load(settings)
YOLO.loadWeights!(model, settings)

voc = YOLO.datasets.VOC.populate()
vocloaded = YOLO.load(voc, settings, indexes = [100]) #load image #100 (a single image)

#Run the model
res = model(vocloaded.imstack_mat);

#Convert the output into readable predictions
predictions = YOLO.postprocess(res, settings, conf_thresh = 0.3, iou_thresh = 0.3)

Testing a single custom image

To pass an image through, the image needs to be loaded, and scaled to the appropriate input size. For YOLOv2-tiny that would be (w, h, color_channels, minibatch_size) == (416, 416, 3, 1).

loadResizePadImageToFit can be used to load, resize & pad the image, while maintaining aspect ratio and anti-aliasing during the resize process.

using YOLO
## Load once
settings = YOLO.pretrained.v2_tiny_voc.load(minibatch_size=1) #run 1 image at a time
model = YOLO.v2_tiny.load(settings)
YOLO.loadWeights!(model, settings)

## Run for each image
imgmat = YOLO.loadResizePadImageToFit("image.jpeg", settings)
res = model(imgmat)
predictions = YOLO.postprocess(res, settings, conf_thresh = 0.3, iou_thresh = 0.3)

or if the image is already in memory

imgmat = loadResizePadImageToFit(img, settings)
res = model(imgmat)
predictions = YOLO.postprocess(res, settings, conf_thresh = 0.3, iou_thresh = 0.3)

Rendering results

To render results, first load Makie before YOLO (in a fresh julia instance):

using Makie, YOLO
## Repeat all above steps to load & run the model
scene = YOLO.renderResult(vocloaded.imstack_mat[:,:,:,1], predictions, settings, save_file = "test.png")
display(scene)

Testing inference speed

The package tests include a small benchmark. A 2018 macbook pro i7. CPU-only:

[ Info: YOLO_v2_tiny inference time per image: 0.1313 seconds (7.62 fps)
[ Info: YOLO_v2_tiny postprocess time per image: 0.0023 seconds (444.07 fps)
[ Info: Total time per image: 0.1336 seconds (7.49 fps)

An i7 desktop with a GTX 1070 GPU:

[ Info: YOLO_v2_tiny inference time per image: 0.0039 seconds (254.79 fps)
[ Info: YOLO_v2_tiny postprocess time per image: 0.0024 seconds (425.51 fps)
[ Info: Total time per image: 0.0063 seconds (159.36 fps)