COCO dataset limited to person class.
| Subset | Quantity |
|---|---|
| Training | 2000 |
| Validation | 250 |
| Testing | 250 |
Test set is a part of the original train set, as there is no access to original test set labels.
Run the initialization script to download and prepare the COCO dataset:
python src/init.pyThis script will download the COCO dataset (limited to the person class) and organize it into training, validation, and testing subsets.
To visualize and explore the dataset using the FiftyOne GUI:
python -m fiftyone app launch dataset_nameOr, to use FiftyOne within a Jupyter notebook:
import fiftyone as fo
# Load your dataset
dataset = fo.load_dataset("dataset_name")
# Launch the app
session = fo.launch_app(dataset)