Using YoloEase in EyeAuras

YoloEase prepares the .onnx model. EyeAuras uses it in ML Search, and then the Behavior Tree decides what to do with the detected objects.

Keep the .pt file for training and fine-tuning. For ML Search, use the .onnx file.

For a full walkthrough, see EyeAuras integration.

Quick start

1. Open the output folder from Trainer.
2. Take the .onnx model.
3. In EyeAuras, open ML Search.
4. Click the model load button and select the .onnx file.
5. Set the capture area so it matches the data used during training.
6. If you trained the model on frames with EyeAuras effects enabled, turn on the same effects before detection.
7. Enable Enable OSD and verify the bounding boxes without clicking anything.
8. Connect ML Find Class, then Mouse Move, and only add Key Press after everything works correctly.

Ready-made example

• Recorded session
• Ready-made EyeAuras pack
• ONNX weights

Read next

• EyeAuras integration
• YOLO ONNX and model weights
• Diagnostics
• AI Search Trigger
• Behavior Trees