Long-tail object detection suffers from poor performance
on tail categories. We reveal that the real culprit lies in the
extremely imbalanced distribution of the classifier’s weight
norm. For conventional softmax cross-entropy loss, such
imbalanced weight norm distribution yields ill conditioned
decision boundary for categories which have small weight
norms. To get rid of this situation, we choose to maximize the cosine similarity between the learned feature and
the weight vector of target category rather than the innerproduct of them. The decision boundary between any two