Hi all,
I am using YOLO 26 to detect people and animals, but it is just the standard model and has not been retrained.
I would like to hear from others who have experience working with the Hailo-8L and Raspberry Pi. Which model works best with this module when used with a Raspberry Pi?
My current understanding is that it is best to use the most recent version of YOLO, but I would be interested to learn how others are using these models for detection.
I recommend taking a look at the Model Explorer in the Developer Zone. Choosing the right model depends on several factors:
- FPS (performance): Ensure the model can process your workload at the required speed. This primarily depends on the number of cameras and the frame rate of each stream.
- Accuracy: In most cases, you should choose the most accurate model that still meets your required FPS.
- License: For commercial projects, the licensing terms of each model may determine whether a particular model is suitable for your application.
Retraining models with a specific dataset will have an impact on the accuracy of the model and is often preferable over a model trained with a generic dataset.