metadata

library_name: unity-sentis
pipeline_tag: object-detection

YOLOv8n validated for Unity Sentis (Version 2.1.2)

YOLOv8n is a real-time multi-object recognition model confirmed to run in Unity 6000.

How to Use

Create a new scene in Unity 6000;
Install com.unity.sentis version 2.1.2 from the package manager;
Add the RunYOLO8n.cs script to the Main Camera;
Drag the Models/yolov8n.onnx file into the Model Asset field;
Drag the classes.txt file into the Classes Asset field;
Create a GameObject > UI > Raw Image object in the scene, set its width and height to 640, and link it as the Display Image field;
Drag the Border Texture.png file into the Border Texture field;
Select an appropriate font in the Font field;
Put a video file in the Assets/StreamingAssets folder and set the Video Filename field to the filename of the video.

If working correctly you should see something like this:

The NMS selection will be improved in later versions of Sentis. Currently uses singular-class approach.

Unity Sentis is the inference engine that runs in Unity 3D. More information can be found at here

The YOLO models use the GPLv3 license.