lcwlcw123/bkchallenge

CBTNet

We believe our results can be reproduced, but if there is any problem please contact me

Test Results

Download the test results from Google Drive.

Get Data and models

Download the pretrained model from Google Drive, and place it in the folder saved_models. Run the following code to generate test results.

Download the test dataset from Google Drive, and place it in the folder test_data. Run the following code to generate test results.

BKchallenge
│   README.md 
│
└───saved_models
│   │   boken_net_g_29500.pth
│   │   deboken_net_g_2000.pth
│   │   DIS_model.pth
│   │   model.pt
│
│   
└───test_data
    │   00000.src.jpg
    │   00001.src.jpg
    │   00002.src.jpg
    |   .... 

Usage

cd BKchallenge
sh test.sh

Reuslt

the results can be seen in the folder result.

Environment

you can look for the 'requirements.txt' to see the requirements

for a new environment you can run below codes

conda create --name boken_test python==3.8.15

conda activate boken_test

conda install pytorch==1.7.0 torchvision==0.8.0 torchaudio==0.7.0 cudatoolkit=10.1 -c pytorch

pip install requirements.txt

bkchallenge's Issues

Request for Meta.txt File in Test dataset

Is it possible to access the meta.txt file for the testing dataset? Without it, the test set seems incomplete for use.

Additionally, has the competition organizer released the Ground Truth (GT) for the test set post-competition? The testing services are currently closed, making it difficult to access.

Appreciate any assistance on this matter! Thank you very much！