Learning to self-train for semi-supervised few-shot classification

Few-shot classification (FSC) is challenging due to the scarcity of labeled training data (e.g. only one labeled data point per class). Meta-learning has shown to achieve promising results by learning to initialize a classification model for FSC. In this paper we propose a novel semi-supervised meta...

Full description

Saved in:
Bibliographic Details
Main Authors: LI, Xinzhe, SUN, Qianru, LIU, Yaoyao, ZHENG, Shibao, ZHOU, Qin, CHUA, Tat-Seng, SCHIELE, Bernt
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2019
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/4445
https://ink.library.smu.edu.sg/context/sis_research/article/5448/viewcontent/NeurIPS_2019_semi_supervised_camera_ready.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-5448
record_format dspace
spelling sg-smu-ink.sis_research-54482021-02-19T03:10:44Z Learning to self-train for semi-supervised few-shot classification LI, Xinzhe SUN, Qianru LIU, Yaoyao ZHENG, Shibao ZHOU, Qin CHUA, Tat-Seng SCHIELE, Bernt Few-shot classification (FSC) is challenging due to the scarcity of labeled training data (e.g. only one labeled data point per class). Meta-learning has shown to achieve promising results by learning to initialize a classification model for FSC. In this paper we propose a novel semi-supervised meta-learning method called learning to self-train (LST) that leverages unlabeled data and specifically meta-learns how to cherry-pick and label such unsupervised data to further improve performance. To this end, we train the LST model through a large number of semi-supervised few-shot tasks. On each task, we train a few-shot model to predict pseudo labels for unlabeled data, and then iterate the self-training steps on labeled and pseudo-labeled data with each step followed by fine-tuning. We additionally learn a soft weighting network (SWN) to optimize the self-training weights of pseudo labels so that better ones can contribute more to gradient descent optimization. We evaluate our LST method on two ImageNet benchmarks for semi-supervised few-shot classification and achieve large improvements over the state-of-the-art. 2019-12-01T08:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/4445 https://ink.library.smu.edu.sg/context/sis_research/article/5448/viewcontent/NeurIPS_2019_semi_supervised_camera_ready.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Few-shot learning semi-supervised learning meta-learning image classification Artificial Intelligence and Robotics Computer Sciences Numerical Analysis and Scientific Computing
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Few-shot learning
semi-supervised learning
meta-learning
image classification
Artificial Intelligence and Robotics
Computer Sciences
Numerical Analysis and Scientific Computing
spellingShingle Few-shot learning
semi-supervised learning
meta-learning
image classification
Artificial Intelligence and Robotics
Computer Sciences
Numerical Analysis and Scientific Computing
LI, Xinzhe
SUN, Qianru
LIU, Yaoyao
ZHENG, Shibao
ZHOU, Qin
CHUA, Tat-Seng
SCHIELE, Bernt
Learning to self-train for semi-supervised few-shot classification
description Few-shot classification (FSC) is challenging due to the scarcity of labeled training data (e.g. only one labeled data point per class). Meta-learning has shown to achieve promising results by learning to initialize a classification model for FSC. In this paper we propose a novel semi-supervised meta-learning method called learning to self-train (LST) that leverages unlabeled data and specifically meta-learns how to cherry-pick and label such unsupervised data to further improve performance. To this end, we train the LST model through a large number of semi-supervised few-shot tasks. On each task, we train a few-shot model to predict pseudo labels for unlabeled data, and then iterate the self-training steps on labeled and pseudo-labeled data with each step followed by fine-tuning. We additionally learn a soft weighting network (SWN) to optimize the self-training weights of pseudo labels so that better ones can contribute more to gradient descent optimization. We evaluate our LST method on two ImageNet benchmarks for semi-supervised few-shot classification and achieve large improvements over the state-of-the-art.
format text
author LI, Xinzhe
SUN, Qianru
LIU, Yaoyao
ZHENG, Shibao
ZHOU, Qin
CHUA, Tat-Seng
SCHIELE, Bernt
author_facet LI, Xinzhe
SUN, Qianru
LIU, Yaoyao
ZHENG, Shibao
ZHOU, Qin
CHUA, Tat-Seng
SCHIELE, Bernt
author_sort LI, Xinzhe
title Learning to self-train for semi-supervised few-shot classification
title_short Learning to self-train for semi-supervised few-shot classification
title_full Learning to self-train for semi-supervised few-shot classification
title_fullStr Learning to self-train for semi-supervised few-shot classification
title_full_unstemmed Learning to self-train for semi-supervised few-shot classification
title_sort learning to self-train for semi-supervised few-shot classification
publisher Institutional Knowledge at Singapore Management University
publishDate 2019
url https://ink.library.smu.edu.sg/sis_research/4445
https://ink.library.smu.edu.sg/context/sis_research/article/5448/viewcontent/NeurIPS_2019_semi_supervised_camera_ready.pdf
_version_ 1770574840068046848