Domain Adaptive Transfer Learning

Jiquan Ngiam, Daiyi Peng, Vijay Vasudevan, Simon Kornblith, Quoc Le, Ruoming Pang

Abstract: Transfer learning is a widely used method to build high performing computer vision models. In this paper, we study the efficacy of transfer learning by examining how the choice of data impacts performance. We find that more pre-training data does not always help, and transfer performance depends on a judicious choice of pre-training data. These findings are important given the continued increase in dataset sizes. We further propose domain adaptive transfer learning, a simple and effective pre-training method using importance weights computed based on the target dataset. Our methods achieve state-of-the-art results on multiple fine-grained classification datasets and are well-suited for use in practice.