Laurent/REVA-QCAV
1
0
Fork 0
mirror of https://github.com/Laurent2916/REVA-QCAV.git synced 2024-09-19 19:45:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

Modified to take any image size (with even width, height > width/2)

Browse source 
Former-commit-id: 2751e6a3df45c1527376a4697d3804d683095d83
This commit is contained in:
milesial 2017-11-30 07:19:52 +01:00
parent 7ea54febec
commit 8b614c3e31
3 changed files with 10 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
README.md
View file

@ -6,6 +6,7 @@ This model scored a [dice coefficient](https://en.wikipedia.org/wiki/S%C3%B8rens
The model used for the last submission is stored in the `MODEL.pth` file, if you wish to play with it. The data is available on the [Kaggle website](https://www.kaggle.com/c/carvana-image-masking-challenge/data). The model used for the last submission is stored in the `MODEL.pth` file, if you wish to play with it. The data is available on the [Kaggle website](https://www.kaggle.com/c/carvana-image-masking-challenge/data).
## Usage ## Usage
### Prediction ### Prediction
You can easily test the output masks on your images via the CLI. You can easily test the output masks on your images via the CLI.
@ -13,13 +14,16 @@ To see all options:
`python predict.py -h` `python predict.py -h`
To predict a single image and save it: To predict a single image and save it:
`python predict.py -i image.jpg -o ouput.jpg
`python predict.py -i image.jpg -o ouput.jpg`
To predict a multiple images and show them without saving them: To predict a multiple images and show them without saving them:
`python predict.py -i image1.jpg image2.jpg --viz --no-save` `python predict.py -i image1.jpg image2.jpg --viz --no-save`
You can use the cpu-only version with `--cpu`. You can use the cpu-only version with `--cpu`.
You can specify which model file to use with `--model MODEL.pth`. You can specify which model file to use with `--model MODEL.pth`.
## Note ## Warning
The code and the overall project architecture is a big mess for now, as I left it abandoned when the challenge finished. I will clean it Soon<sup>TM</sup>. In order to process the image, it is splitted into two squares (a left on and a right one), and each square is passed into the net. The two square masks are then merged again to produce the final image. As a consequence, the height of the image must be strictly superior than half the width. Make sure the width is even too.

3
predict.py
View file

@ -12,6 +12,7 @@ from crf import dense_crf
from unet import UNet from unet import UNet
def predict_img(net, full_img, gpu=False): def predict_img(net, full_img, gpu=False):
img = resize_and_crop(full_img) img = resize_and_crop(full_img)
@ -39,7 +40,7 @@ def predict_img(net, full_img, gpu=False):
y_l = F.upsample_bilinear(y_l, scale_factor=2).data[0][0].cpu().numpy() y_l = F.upsample_bilinear(y_l, scale_factor=2).data[0][0].cpu().numpy()
y_r = F.upsample_bilinear(y_r, scale_factor=2).data[0][0].cpu().numpy() y_r = F.upsample_bilinear(y_r, scale_factor=2).data[0][0].cpu().numpy()
y = merge_masks(y_l, y_r, 1918) y = merge_masks(y_l, y_r, full_img.size[0])
yy = dense_crf(np.array(full_img).astype(np.uint8), y) yy = dense_crf(np.array(full_img).astype(np.uint8), y)
return yy > 0.5 return yy > 0.5

1
submit.py
View file

@ -1,3 +1,4 @@
# used to predict all test images and encode results in a csv file
import os import os
from PIL import Image from PIL import Image