Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification

Marzi Heidari, Abdullah Alchihabi, Qing En, Yuhong Guo
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:1369-1377, 2024.

Abstract

Cross-domain few-shot classification induces a much more challenging problem than its in-domain counterpart due to the existence of domain shifts between the training and test tasks. In this paper, we develop a novel Adaptive Parametric Prototype Learning (APPL) method under the meta-learning convention for cross-domain few-shot classification. Different from existing prototypical few-shot methods that use the averages of support instances to calculate the class prototypes, we propose to learn class prototypes from the concatenated features of the support set in a parametric fashion and meta-learn the model by enforcing prototype-based regularization on the query set. In addition, we fine-tune the model in the target domain in a transductive manner using a weighted-moving-average self-training approach on the query instances. We conduct experiments on multiple cross-domain few-shot benchmark datasets. The empirical results demonstrate that APPL yields superior performance to many state-of-the-art cross-domain few-shot learning methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-heidari24a, title = {Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification}, author = {Heidari, Marzi and Alchihabi, Abdullah and En, Qing and Guo, Yuhong}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {1369--1377}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/heidari24a/heidari24a.pdf}, url = {https://proceedings.mlr.press/v238/heidari24a.html}, abstract = {Cross-domain few-shot classification induces a much more challenging problem than its in-domain counterpart due to the existence of domain shifts between the training and test tasks. In this paper, we develop a novel Adaptive Parametric Prototype Learning (APPL) method under the meta-learning convention for cross-domain few-shot classification. Different from existing prototypical few-shot methods that use the averages of support instances to calculate the class prototypes, we propose to learn class prototypes from the concatenated features of the support set in a parametric fashion and meta-learn the model by enforcing prototype-based regularization on the query set. In addition, we fine-tune the model in the target domain in a transductive manner using a weighted-moving-average self-training approach on the query instances. We conduct experiments on multiple cross-domain few-shot benchmark datasets. The empirical results demonstrate that APPL yields superior performance to many state-of-the-art cross-domain few-shot learning methods.} }
Endnote
%0 Conference Paper %T Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification %A Marzi Heidari %A Abdullah Alchihabi %A Qing En %A Yuhong Guo %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-heidari24a %I PMLR %P 1369--1377 %U https://proceedings.mlr.press/v238/heidari24a.html %V 238 %X Cross-domain few-shot classification induces a much more challenging problem than its in-domain counterpart due to the existence of domain shifts between the training and test tasks. In this paper, we develop a novel Adaptive Parametric Prototype Learning (APPL) method under the meta-learning convention for cross-domain few-shot classification. Different from existing prototypical few-shot methods that use the averages of support instances to calculate the class prototypes, we propose to learn class prototypes from the concatenated features of the support set in a parametric fashion and meta-learn the model by enforcing prototype-based regularization on the query set. In addition, we fine-tune the model in the target domain in a transductive manner using a weighted-moving-average self-training approach on the query instances. We conduct experiments on multiple cross-domain few-shot benchmark datasets. The empirical results demonstrate that APPL yields superior performance to many state-of-the-art cross-domain few-shot learning methods.
APA
Heidari, M., Alchihabi, A., En, Q. & Guo, Y.. (2024). Adaptive Parametric Prototype Learning for Cross-Domain Few-Shot Classification. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:1369-1377 Available from https://proceedings.mlr.press/v238/heidari24a.html.

Related Material