MGDNF Datasets

This repository contains the datasets utilized in the MGDNF paper, encompassing the following:

• Yak Individual Category Dataset

• Vehicle Identification Dataset

• Bird Species Dataset

Each dataset is organized into training and testing sets, structured as follows:

Yak Individual Category Dataset

• Description: Comprises images of 21 distinct yak individuals, each representing a unique category.
• Image Count per Category: Ranges from 140 to 200 images.
• Total Images:
  • Training Set: 2,663 images
  • Test Set: 1,213 images
• Structure:
  • yak/
    • train/
      • individual_01/
      • individual_02/
      • ...
    • test/
      • individual_01/
      • individual_02/
      • ...

Vehicle Identification Dataset

• Description: Contains images of 10 different vehicle types.
• Total Images:
  • Training Set: 2,200 images
  • Test Set: 1,000 images
• Structure:
  • vehicle_identification/
    • train/
      • type_01/
      • type_02/
      • ...
    • test/
      • type_01/
      • type_02/
      • ...

Bird Species Dataset

• Description: Features images of 24 bird species.
• Total Images: 5,456
  • Training Set: 3,868 images
  • Test Set: 1,588 images
• Structure:
  • bird_species/
    • train/
      • species_01/
      • species_02/
      • ...
    • test/
      • species_01/
      • species_02/
      • ...

Usage

1. Cloning the Repository:

   git clone https://github.com/Deep-AI-Application-DAIP/MGDNF-datasets.git

2. Accessing Datasets:

   Navigate to the respective dataset directories (yak/, vehicle_identification/, bird_species/) to access the training and testing images.

Citation

If you utilize these datasets in your research, please cite the MGDNF paper accordingly.

License

These datasets are provided for academic and research purposes. For other uses, please contact the repository maintainers.