> For the complete documentation index, see [llms.txt](https://docs.roboflow.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.roboflow.com/datasets/versions.md).

# Versions

- [Dataset Versions](https://docs.roboflow.com/datasets/versions/dataset-versions.md): Create dataset versions as point-in-time snapshots with preprocessing and augmentation before training.
- [Create a Dataset Version](https://docs.roboflow.com/datasets/versions/dataset-versions/create-a-dataset-version.md): Create a dataset version for use in training a model.
- [Delete a Version](https://docs.roboflow.com/datasets/versions/dataset-versions/delete-a-version.md): Move a dataset version to Trash, where it is kept for 30 days before being permanently removed.
- [Export a Dataset Version](https://docs.roboflow.com/datasets/versions/dataset-versions/exporting-data.md): Export data from Roboflow for training.
- [Preprocess Images](https://docs.roboflow.com/datasets/versions/dataset-versions/image-preprocessing.md): Image preprocessing steps to prepare data for models.
- [Image Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation.md): Create augmented images to improve model performance.
- [Add Augmentations to Images](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/add-augmentations-to-images.md): Add augmentation steps when creating a dataset version to generate additional training images.
- [Augmentation Types](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types.md): Overview of the image augmentation types Roboflow offers and how each transforms your dataset.
- [90º Rotate Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/90o-rotate-augmentation.md): Randomly rotate images 90 or 180 degrees so your model is insensitive to camera orientation.
- [Blur Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/blur-augmentation.md): Add random Gaussian blur to make your model more resilient to variations in camera focus.
- [Brightness Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/brightness-augmentation.md): Randomly brighten or darken images to make your model resilient to lighting and camera setting changes.
- [Camera Gain Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/camera-gain-augmentation.md): Simulate camera sensor gain (ISO) noise to make your model resilient to sensor and lighting variation.
- [Crop Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/crop-augmentation.md): Randomly crop a subset of each image to add positional and scale variability and improve model generalization.
- [Cutout Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/cutout-augmentation.md): Randomly mask portions of an image with black boxes to make your model resilient to object occlusion.
- [Exposure Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/exposure-augmentation.md): Adjust gamma exposure to brighten or darken images and build resilience to lighting and camera changes.
- [Flip Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/flip-augmentation.md): Randomly flip images horizontally or vertically so your model is insensitive to subject orientation.
- [Grayscale Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/grayscale-augmentation.md): Probabilistically convert a percentage of your dataset's images to grayscale so your model relies less on color.
- [Hue Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/hue-augmentation.md): Randomly shift image hues around the color wheel to make your model resilient to color variation.
- [Mosaic Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/mosaic-augmentation.md): Combine multiple images into a single tile to help your model perform better on small objects.
- [Motion Blur Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/motion-blur-augmentation.md): Simulate camera or subject motion with directional blur to make your model resilient to movement.
- [Noise Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/noise-augmentation.md): Inject random salt-and-pepper noise into images to make your model resilient to camera artifacts.
- [Rotation Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/rotation-augmentation.md): Randomly rotate images clockwise or counter-clockwise up to a chosen degree to handle camera roll.
- [Saturation Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/saturation-augmentation.md): Randomly adjust color vibrancy in images to make your model resilient to saturation differences.
- [Shear Augmentation](https://docs.roboflow.com/datasets/versions/dataset-versions/image-augmentation/augmentation-types/shear-augmentation.md): Randomly shear images along their horizontal or vertical axis to add perspective variability.
- [Dataset Health Check](https://docs.roboflow.com/datasets/versions/dataset-health-check.md): Assess and improve the quality of your dataset.