In realistic scenarios, the effectiveness of Deep Neural Networks is hindered by domain shift,
where discrepancies between training (source) and testing (target) domains lead to poor …