TGDM: Target Guided Dynamic Mixup for cross-domain few-shot learning

• Main Authors: ZHUO, Linhai, FU, Yuqian, CHEN, Jingjing, CAO, Yixin, JIANG, Yu-Gang
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2022
• Subjects: cross-domain few-shot learning; dynamic mixup; target guided learning; bi-level meta-learning; Databases and Information Systems; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/7453; https://ink.library.smu.edu.sg/context/sis_research/article/8456/viewcontent/3503161.3548052.pdf
• Institution: Singapore Management University

Given sufficient training data on the source domain, cross-domain few-shot learning (CD-FSL) aims at recognizing new classes with a small number of labeled examples on the target domain. The key to addressing CD-FSL is to narrow the domain gap and transferring knowledge of a network trained on the source domain to the target domain. To help knowledge transfer, this paper introduces an intermediate domain generated by mixing images in the source and the target domain. Specifically, to generate the optimal intermediate domain for different target data, we propose a novel target guided dynamic mixup (TGDM) framework that leverages the target data to guide the generation of mixed images via dynamic mixup. The proposed TGDM framework contains a Mixup-3T network for learning classifiers and a dynamic ratio generation network (DRGN) for learning the optimal mix ratio. To better transfer the knowledge, the proposed Mixup-3T network contains three branches with shared parameters for classifying classes in the source domain, target domain, and intermediate domain. To generate the optimal intermediate domain, the DRGN learns to generate an optimal mix ratio according to the performance on auxiliary target data. Then, the whole TGDM framework is trained via bi-level meta-learning so that TGDM can rectify itself to achieve optimal performance on target data. Extensive experimental results on several benchmark datasets verify the effectiveness of our method