---
pipeline_tag: object-detection
license: apache-2.0
library_name: unity-sentis
---

# Iris Landmark from Google Mediapipe validated for Unity Sentis (Version 1.4.0-pre.2*)
*Version 1.3.0 Sentis files are not compatible with Sentis 1.4.0 and need to be recreated/downloaded

This is the [Iris Landmark model](https://github.com/google/mediapipe/blob/master/docs/solutions/iris.md) from Google in the Sentis format.

The model detects 3D markers on and around a centered image of an eye. You could use these markers, to detect eye movements.

**IMPORTANT:** The eye needs to be centered and cropped to fit the image showing a region around the eye. For images with faces not in the center or for multiple faces, you will need another model to detect faces, and crop them before feeding them into this model. For example you could use [Blaze Face](https://huggingface.co/unity/sentis-blaze-face) to detect them
Then you can use the [face landmark](https://huggingface.co/unity/sentis-face-landmarks) model to single out the eye region to send to this model.

## How to Use
* Create a new scene in Unity 2023
* Install `com.unity.sentis` version `1.4.0-pre.3` from the package manager
* Put the iris_landmark.sentis file in the `Assets/StreamingAssets` folder
* Put a video in the `Assets/StreamingAssets` folder and set `videoName` variable to the video name
* Create a RawImage and place it in your scene. Link to this image in the `previewUI` field.

## Preview
If you get it working it should look like this (original image from pexels.com):
![image showing markers](preview_iris_landmark.png)

## License
All Google Mediapipe models are open source under the Apache 2.0 license. The accompanying C# source code we provide can be used in your applications for commercial purposes.