Run in Google Colab

View source on GitHub

Download notebook

# Exploring the Dataset Zoo This experience introduces you to the core components of the FiftyOne Zoo: - The **Dataset Zoo** for accessing and exploring public datasets - The **Model Zoo** for running pre-trained models on your data - Creating your **own remotely-sourced datasets** for reuse and collaboration Whether you’re a researcher, engineer, or educator, these tools help streamline your computer vision workflows in FiftyOne. > 💡 Make sure to run `pip install fiftyone torch torchvision` before starting. ## FiftyOne Zoo: A Hub for Datasets and Models FiftyOne Zoo provides easy access to a vast collection of pre-built datasets and pre-trained models. This notebook will guide you through exploring and using these resources. ### Key Components: - **Dataset Zoo:** Offers a wide range of computer vision datasets, ready for immediate use. - **Model Zoo:** Provides pre-trained models for various tasks, enabling quick experimentation and deployment. Let’s dive in! ## Dataset Zoo ### Exploring the Dataset Zoo The Dataset Zoo simplifies the process of loading and working with popular datasets. #### Listing Available Datasets ### Loading a Dataset (Example: MNIST) ### Visualizing the Dataset  ### Loading a Specific Split (Example: COCO) ### Downloading and Loading a Dataset with Specific Splits and Downsampling (Example: open-images-v6) ### Working with Dataset Metadata ### Example: Loading a Remote Image Dataset With fiftyOne you can work/create zoo datasets whose download/preparation methods are hosted via GitHub repositories or public URLs ### Other loading examples with remote datasets Load 50 random samples from the validation split Only the required images will be downloaded (if necessary). By default, only detections are loaded Load segmentations for 25 samples from the validation split that contain cats and dogs Images that contain all `classes` will be prioritized first, followed by images that contain at least one of the required `classes`. If there are not enough images matching `classes` in the split to meet `max_samples`, only the available images will be loaded. Images will only be downloaded if necessary Download the entire validation split and load both detections and segmentations. Subsequent partial loads of the validation split will never require downloading any images.