/GeoLifeCLEF2024-2nd-place

[GeoLifeCLEF 2024] Using Samples with Single-label to train hybrid model for Species Distribution Modeling

Primary LanguageJupyter Notebook

GeoLifeCLEF2024-2nd-place

[GeoLifeCLEF 2024] Using Samples with Single-label to train hybrid model for Species Distribution Modeling

Components

  1. train_and_make_pseudo.ipynb

    • Trains the initial model using original labels
    • Uses the trained model to perform pseudo-labeling on PO data
  2. train_with_pseudo_label.ipynb

    • Trains a new model using both original labels and pseudo labels

Detailed Information

For a comprehensive explanation of our methodology and results, please refer to our research paper (link to paper).