Few-Shot Adversarial Domain Adaptation
This addresses the problem of domain adaptation for machine learning practitioners when labeled data is scarce, offering a novel method that is incremental in improving few-shot performance.
The paper tackles supervised domain adaptation with few labeled target samples by using an adversarial learning framework that aligns semantic embeddings while maximizing domain confusion, achieving effective adaptation with as few as one labeled sample per category.
This work provides a framework for addressing the problem of supervised domain adaptation with deep models. The main idea is to exploit adversarial learning to learn an embedded subspace that simultaneously maximizes the confusion between two domains while semantically aligning their embedding. The supervised setting becomes attractive especially when there are only a few target data samples that need to be labeled. In this few-shot learning scenario, alignment and separation of semantic probability distributions is difficult because of the lack of data. We found that by carefully designing a training scheme whereby the typical binary adversarial discriminator is augmented to distinguish between four different classes, it is possible to effectively address the supervised adaptation problem. In addition, the approach has a high speed of adaptation, i.e. it requires an extremely low number of labeled target training samples, even one per category can be effective. We then extensively compare this approach to the state of the art in domain adaptation in two experiments: one using datasets for handwritten digit recognition, and one using datasets for visual object recognition.