Transfer learning is a critical technique in training deep neural networks to tackle downstream tasks with better or faster solutions, by leveraging previously acquired knowledge. This is especially ...