Domain Adaptive Transfer Learning with Specialist Models
This addresses the problem of inefficient transfer learning for computer vision practitioners by optimizing pre-training data choice, though it is incremental as it builds on existing domain adaptation ideas.
The study investigated how pre-training data selection affects transfer learning performance, finding that more data does not always improve results, and proposed a domain adaptive transfer learning method using importance weights based on the target dataset, achieving state-of-the-art results on multiple fine-grained classification datasets.
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 method to compute importance weights follow from ideas in domain adaptation, and we show a novel application to transfer learning. Our methods achieve state-of-the-art results on multiple fine-grained classification datasets and are well-suited for use in practice.