Competition URL: https://tbrain.trendmicro.com.tw/Competitions/Details/20
Private Leaderboard: 14 / 743 (Top 2%)

• Clone this repo to your local

git clone https://github.com/come880412/Orchid219_classification
cd Orchid219_classification

• System: Ubuntu20.04

• Pytorch version: Pytorch 1.7 or higher

• Python version: Python 3.7

• Testing:
  CPU: AMR Ryzen 7 4800H with Radeon Graphics RAM: 32GB
  GPU: NVIDIA GeForce RTX 1660Ti 6GB

• Training:
  CPU: Intel(R) Xeon(R) Gold 6226R CPU @ 2.90GHz RAM: 256GB GPU: NVIDIA GeForce RTX 3090 24GB * 2

Please read the "requirement.txt" for the details.

• You should prepare the dataset from here, and put the dataset on the folder ../dataset/Orchid219. After doing so, please use the following command to do data preprocessing.

python3 preprocessing.py 

• Note: please modify the dataset path on the script preprocessing.py.

• Please download the pretrained models from here, and put the models on the folder ./public_model.

Orchid
├── Orchid219_classification/ 
├── dataset/ 
  ├── Orchid219
    ├── images
    ├── private_and_public
      ├── public images and private images
    ├── submission_template.csv
    ├── label.csv

If you would like to use colab to implement, please use the file "Orchid.ipynb".

• After preparing the dataset and downloading the pretrained models, you could use the following command to generate the .csv file, which is the best public score on our submission.

bash inference.sh

• In this competition, we use three models, including swin_large, beit_large_384, convnext_xlarge [1]. You could check the training detail on the script "train.py", and train all models using the following command.

bash train.sh

[1] https://github.com/rwightman/pytorch-image-models