Why do we set a custom bias for classifier head in Object Detection?

(Jayesh Saita) #1

I am studying Object detection and was going through the 2018 part 2 SSD notebook and the new Retinanet notebook. Both implementations set a custom bias for classifier head of object detection. In SSD, the bias is -3, and in Retinanet, it is -4.
SSD



Retinanet

retinanet%20bias%20call

What is the reason for setting this custom bias for classifier head? Also, is there any specific reason behind the numbers -3 and -4?

0 Likes

(Karl) #2

Bias initialization comes from the Focal Loss paper

2 Likes