petrosgk / Kaggle Carvana Image Masking Challenge
Licence: mit
Stars: ✭ 256
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Kaggle Carvana Image Masking Challenge
Pytorch Toolbelt
PyTorch extensions for fast R&D prototyping and Kaggle farming
Stars: ✭ 942 (+267.97%)
Mutual labels: kaggle, image-segmentation
Image Segmentation
Implementation of FCN (8/16/32) in Tensorflow. 🌀 在TensorFlow框架下实现 FCN (全卷积神经网络) 。
Stars: ✭ 19 (-92.58%)
Mutual labels: image-segmentation
Kaggle Competition
Summary of the Kaggle Stock Prediction Competition & my Trial
Stars: ✭ 78 (-69.53%)
Mutual labels: kaggle
HyperDenseNet pytorch
Pytorch version of the HyperDenseNet deep neural network for multi-modal image segmentation
Stars: ✭ 58 (-77.34%)
Mutual labels: image-segmentation
ashrae-great-energy-predictor-3-solution-analysis
Analysis of top give winning solutions of the ASHRAE Great Energy Predictor III competition
Stars: ✭ 44 (-82.81%)
Mutual labels: kaggle
kernel-run
Run any Jupyter notebook instantly using Kaggle kernels
Stars: ✭ 59 (-76.95%)
Mutual labels: kaggle
coursera-ai-for-medicine-specialization
Programming assignments, labs and quizzes from all courses in the Coursera AI for Medicine Specialization offered by deeplearning.ai
Stars: ✭ 80 (-68.75%)
Mutual labels: image-segmentation
ResUNetPlusPlus
Official code for ResUNetplusplus for medical image segmentation (TensorFlow implementation) (IEEE ISM)
Stars: ✭ 69 (-73.05%)
Mutual labels: image-segmentation
TensorFlow-Advanced-Segmentation-Models
A Python Library for High-Level Semantic Segmentation Models based on TensorFlow and Keras with pretrained backbones.
Stars: ✭ 64 (-75%)
Mutual labels: image-segmentation
HumanOrRobot
a solution for competition of kaggle `Human or Robot`
Stars: ✭ 16 (-93.75%)
Mutual labels: kaggle
game2vec
TensorFlow implementation of word2vec applied on https://www.kaggle.com/tamber/steam-video-games dataset, using both CBOW and Skip-gram.
Stars: ✭ 62 (-75.78%)
Mutual labels: kaggle
Reproducibilty-Challenge-ECANET
Unofficial Implementation of ECANets (CVPR 2020) for the Reproducibility Challenge 2020.
Stars: ✭ 27 (-89.45%)
Mutual labels: image-segmentation
COVID-19-CaseStudy-and-Predictions
This repository is a case study, analysis and visualization of COVID-19 Pandemic spread along with prediction models.
Stars: ✭ 90 (-64.84%)
Mutual labels: kaggle
DLCV2018SPRING
Deep Learning for Computer Vision (CommE 5052) in NTU
Stars: ✭ 38 (-85.16%)
Mutual labels: image-segmentation
convolutedPredictions Cdiscount
2nd place solution to Kaggle's Cdiscount image classification challange.
Stars: ✭ 17 (-93.36%)
Mutual labels: kaggle
Image-Segmentation-Loss-Functions
some loss functions of image segmentation
Stars: ✭ 56 (-78.12%)
Mutual labels: image-segmentation
Kaggle Carvana Image Masking Challenge solution with Keras
This solution was based on Heng CherKeng's code for PyTorch. I kindly thank him for sharing his work. 128x128, 256x256, 512x512 and 1024x1024 U-nets are implemented. Public LB scores for each U-net are:
U-net | LB score |
---|---|
128x128 | 0.990 |
256x256 | 0.992 |
512x512 | 0.995 |
1024x1024 | 0.996 |
Updates
Update 28.8.2017
- Added loss with weighted boundary (thanks to lyakaap)
Update 15.8.2017
- Added Hue/Saturation/Value augmentation.
- Switched to RMSprop optimizer as default.
- Added multithreaded inference with inference and data loading done on separate threads. This reduced inference time by 40% in my tests. You can run
test_submit_multithreaded.py
to try it.
Update 10.8.2017
- Added 1024x1024 U-net
- Not using predict_generator anymore due to memory constraints with large input
Update 9.8.2017
- Using Binary Crossentropy Dice Loss in place of Binary Crossentropy
- Callbacks now use val_dice_loss as a metric in place of val_loss
Requirements
- Keras 2.0 w/ TF backend
- sklearn
- cv2
- tqdm
- h5py
Usage
Data
Place 'train', 'train_masks' and 'test' data folders in the 'input' folder.
Convert training masks to .png format. You can do this with:
mogrify -format png *.gif
in the 'train_masks' data folder.
Train
Run python train.py
to train the model.
Test and submit
Run python test_submit.py
to make predictions on test data and generate submission.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].