jacobkie / 2018dsb Goto Github PK
View Code? Open in Web Editor NEW2018 Data Science Bowl 2nd Place Solution
License: MIT License
2018 Data Science Bowl 2nd Place Solution
License: MIT License
label_mask = remove_small_holes(label_mask)
label_mask = basin(label_mask, wall)
label_mask = remove_small_holes(label_mask)
Are these function working on instance labels or semantic labels
@jacobkie
Can you share links to weights you have trained?
when I'm trying to run predict_auto.py in the folder script_final, there occurs an error:
from keras import Model, optimizers,losses, activations, models
ImportError: cannot import name 'Model'
It may due to the version of keras I installed do not fit the project. But the project didn't provide the required package version(maybe I do not find?)๐ญใ I tried to search in the Internet but still cannot solve the problem...
So could you please tell me what's the version need to run the project? Thank you so much!
when run 'python3 train_ext.py', meet some issues, is my environment not correct?
can you help to provide your running environment? by running the 'pip3 list'.
`
Epoch 1/100
Traceback (most recent call last):
File "train_ext.py", line 62, in
train_1s()
File "train_ext.py", line 59, in train_1s
model.train_generator(tr_ms1, val_ms1, 1e-3, 100, 'all')
File "/2th-DSB2018/script_final/model_rcnn_weight.py", line 495, in train_generator
use_multiprocessing=False,
File "/usr/local/lib/python3.5/dist-packages/keras/legacy/interfaces.py", line 91, in wrapper
return func(*args, **kwargs)
File "/usr/local/lib/python3.5/dist-packages/keras/engine/training.py", line 2145, in fit_generator
generator_output = next(output_generator)
File "/usr/local/lib/python3.5/dist-packages/keras/utils/data_utils.py", line 770, in get
six.reraise(value.class, value, value.traceback)
File "/usr/lib/python3/dist-packages/six.py", line 686, in reraise
raise value
File "/usr/local/lib/python3.5/dist-packages/keras/utils/data_utils.py", line 635, in _data_generator_task
generator_output = next(self._generator)
File "/medical_data/yunhai/2th-DSB2018/script_final/model_ms1.py", line 42, in generator_1s_v11
image, mask = next(gen)
File "train_ext.py", line 34, in train_generator
image, mask = next(gen)
File "/2th-DSB2018/script_final/generator.py", line 64, in data_generator_multi
zoom_range = config.ZOOM_RANGE)
File "/2th-DSB2018/script_final/preprocess.py", line 198, in affine_transform_batch
xt = scipy.ndimage.affine_transform(x, matrix, order=1, cval=-512)
File "/usr/local/lib/python3.5/dist-packages/scipy/ndimage/interpolation.py", line 449, in affine_transform
raise RuntimeError('affine matrix has wrong number of rows')
RuntimeError: affine matrix has wrong number of rows
`
I am trying to use your pre-trained weights to run predict_auto.py, but it complains that there is no hdf5 file in the cache folder. If I run merge_model_weight.sh
in said folder first, then it will load the merged h5 file. But, this process takes a very long time (~30min??) in 25 stages, i.e. the terminal output from the first stage looks like
0/25 2 (205, 694, 3) 704
19-02-22 23:58:35 INFO| loading weights form /home/ubuntu/git_repos/2018DSB/cache/UnetRCNN_180410-221747/71_0.4155.hdf5
Is this what I should be doing and expect to see?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.