Questions about the model

[kevinzezel]

I have an application running standard yolov8, where I only filter the class number 0, person. Does this model with the coco dataset have the same accuracy as yolov8, but because it only has one class it runs faster? Furthermore, I notice many false positives from class person with other classes in the model. In this case, does this model also have fewer false positives?

Regards, Kevin