Skip to content

Latest commit

 

History

History
83 lines (68 loc) · 2.03 KB

README.md

File metadata and controls

83 lines (68 loc) · 2.03 KB
Raw

このリポジトリは第6回Brain(s)コンテスト by FUJIFILM AI Academy Brain(s)におけるQ2:2nd, Q3:1st place, 総合1位🥇の解法です.

demo notebookにsolutionが記述されています.

🔥Demo Notebook(Google CoLab)

requirements

albumentations              1.0.3
black                       21.7b0
flake8                      3.9.2
isort                       5.9.3
matplotlib                  3.4.3
mypy                        0.910
opencv-python               4.5.3.56
Pillow                      8.3.1
pytorch-widedeep            1.0.5
scikit-learn                0.24.2
segmentation-models-pytorch 0.2.0
timm                        0.4.12
torch                       1.8.1+cu111
torchaudio                  0.8.1
torchmetrics                0.5.0
torchvision                 0.9.1+cu111

Checking format

sh check_format.sh DIR SCRIPT ...

e. g.

sh check_format.sh libs train.py train_seg.py

Q1

in ./solution.ipynb

Q2

overview

Q2

training

  • classification
CUDA_VISIBLE_DEVICES=0 python3 train.py config/xxxx.yaml
  • segmentation
CUDA_VISIBLE_DEVICES=0 python3 train_seg.py config/xxxx.yaml

inference

  • classification
CUDA_VISIBLE_DEVICES=0 python3 Q2_classification_inference.py config/xxxx.yaml --top-k 5
  • segmentation
CUDA_VISIBLE_DEVICES=0 python3 segmentation_inference.py config/xxxx.yaml --wall-type W --threshold 0.9

Q3

overview

Q3

training

CUDA_VISIBLE_DEVICES=0 python3 train_seg.py config/xxxx.yaml

inference

CUDA_VISIBLE_DEVICES=0 python3 segmentation_inference.py config/xxxx.yaml --threshold 0.5 --sub-pcon PCON_CONFIG_NAME --save-full-image

zip

python3 pack.py config/xxxx.yaml