Correct score for one-class detection

[makra89]

See comment in See #311

I think the score in yolo_nms is calculated wrong when only having one class. The model cannot learn anything about the class since the associated loss is always 0 --> class probability is always <= 0.5 --> score is always <= 0.5

Ah and by the way: Awesome work :)
Your model works very well when trained trying to build an aimbot: https://github.com/makra89/Quake_AI

[MatthewPerryman]

I don't know if it's commonplace to write on these, but with this change my one class object detector works perfectly thanks!