Adapting a model to perform well on unforeseen data outside its training set is a common problem that continues to motivate new approaches. We demonstrate that application of batch normalization in the output layer, prior to softmax activation, results in improved generalization across visual data domains in a refined ResNet model. The approach adds negligible computational complexity yet outperforms many domain adaptation methods that explicitly learn to align data domains. We benchmark this technique on the Office-Home dataset and show that batch normalization is competitive with other leading methods. We show that this method is not sensitive to presence of source data during adaptation and further present the impact on trained tensor distributions tends toward sparsity. Code is available at https://github.com/matthewbehrend/BNC
Download