Parameterless Transductive Feature Re-representation for Few-Shot Learning

Wentao Cui, Yuhong Guo
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:2212-2221, 2021.

Abstract

Recent literature in few-shot learning (FSL) has shown that transductive methods often outperform their inductive counterparts. However, most transductive solutions, particularly the meta-learning based ones, require inserting trainable parameters on top of some inductive baselines to facilitate transduction. In this paper, we propose a parameterless transductive feature re-representation framework that differs from all existing solutions from the following perspectives. (1) It is widely compatible with existing FSL methods, including meta-learning and fine tuning based models. (2) The framework is simple and introduces no extra training parameters when applied to any architecture. We conduct experiments on three benchmark datasets by applying the framework to both representative meta-learning baselines and state-of-the-art FSL methods. Our framework consistently improves performances in all experiments and refreshes the state-of-the-art FSL results.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-cui21a, title = {Parameterless Transductive Feature Re-representation for Few-Shot Learning}, author = {Cui, Wentao and Guo, Yuhong}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {2212--2221}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/cui21a/cui21a.pdf}, url = {https://proceedings.mlr.press/v139/cui21a.html}, abstract = {Recent literature in few-shot learning (FSL) has shown that transductive methods often outperform their inductive counterparts. However, most transductive solutions, particularly the meta-learning based ones, require inserting trainable parameters on top of some inductive baselines to facilitate transduction. In this paper, we propose a parameterless transductive feature re-representation framework that differs from all existing solutions from the following perspectives. (1) It is widely compatible with existing FSL methods, including meta-learning and fine tuning based models. (2) The framework is simple and introduces no extra training parameters when applied to any architecture. We conduct experiments on three benchmark datasets by applying the framework to both representative meta-learning baselines and state-of-the-art FSL methods. Our framework consistently improves performances in all experiments and refreshes the state-of-the-art FSL results.} }
Endnote
%0 Conference Paper %T Parameterless Transductive Feature Re-representation for Few-Shot Learning %A Wentao Cui %A Yuhong Guo %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-cui21a %I PMLR %P 2212--2221 %U https://proceedings.mlr.press/v139/cui21a.html %V 139 %X Recent literature in few-shot learning (FSL) has shown that transductive methods often outperform their inductive counterparts. However, most transductive solutions, particularly the meta-learning based ones, require inserting trainable parameters on top of some inductive baselines to facilitate transduction. In this paper, we propose a parameterless transductive feature re-representation framework that differs from all existing solutions from the following perspectives. (1) It is widely compatible with existing FSL methods, including meta-learning and fine tuning based models. (2) The framework is simple and introduces no extra training parameters when applied to any architecture. We conduct experiments on three benchmark datasets by applying the framework to both representative meta-learning baselines and state-of-the-art FSL methods. Our framework consistently improves performances in all experiments and refreshes the state-of-the-art FSL results.
APA
Cui, W. & Guo, Y.. (2021). Parameterless Transductive Feature Re-representation for Few-Shot Learning. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:2212-2221 Available from https://proceedings.mlr.press/v139/cui21a.html.

Related Material