> 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/dataset-versions/image-augmentation/augmentation-types/flip-augmentation.md).

# Flip Augmentation

Randomly flip (reflect) an image vertically or horizontally to help your model be insensitive to subject orientation.

* **Horizontal:** Flip the image’s [NumPy array](https://numpy.org/) in the left/right direction.
* **Vertical:** Flip the image’s [NumPy array](https://numpy.org/) in the up/down direction.

## Example

<figure><img src="/files/ufNS2MakqYkLUS9g2SLi" alt=""><figcaption><p>Original Image</p></figcaption></figure>

<figure><img src="/files/rs60pCkof6Xugmv7pmW0" alt=""><figcaption><p>Flip Augmentation Horizontal</p></figcaption></figure>

<figure><img src="/files/MpNSVEUdcLAPad4WVBJv" alt=""><figcaption><p>Flip Augmentation Vertical</p></figcaption></figure>

## Learn More

* [See how Flip Augmentation Improves Model Performance](https://blog.roboflow.com/how-flip-augmentation-improves-model-performance/)