Adaptive Negative Evidential Deep Learning for Open-set Semi-supervised Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Yang, Deng, Danruo, Liu, Furui, Jin, Yueming, Dou, Qi, Chen, Guangyong, Heng, Pheng-Ann
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911837837066240
author Yu, Yang
Deng, Danruo
Liu, Furui
Jin, Yueming
Dou, Qi
Chen, Guangyong
Heng, Pheng-Ann
author_facet Yu, Yang
Deng, Danruo
Liu, Furui
Jin, Yueming
Dou, Qi
Chen, Guangyong
Heng, Pheng-Ann
contents Semi-supervised learning (SSL) methods assume that labeled data, unlabeled data and test data are from the same distribution. Open-set semi-supervised learning (Open-set SSL) considers a more practical scenario, where unlabeled data and test data contain new categories (outliers) not observed in labeled data (inliers). Most previous works focused on outlier detection via binary classifiers, which suffer from insufficient scalability and inability to distinguish different types of uncertainty. In this paper, we propose a novel framework, Adaptive Negative Evidential Deep Learning (ANEDL) to tackle these limitations. Concretely, we first introduce evidential deep learning (EDL) as an outlier detector to quantify different types of uncertainty, and design different uncertainty metrics for self-training and inference. Furthermore, we propose a novel adaptive negative optimization strategy, making EDL more tailored to the unlabeled dataset containing both inliers and outliers. As demonstrated empirically, our proposed method outperforms existing state-of-the-art methods across four datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2303_12091
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Adaptive Negative Evidential Deep Learning for Open-set Semi-supervised Learning
Yu, Yang
Deng, Danruo
Liu, Furui
Jin, Yueming
Dou, Qi
Chen, Guangyong
Heng, Pheng-Ann
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Semi-supervised learning (SSL) methods assume that labeled data, unlabeled data and test data are from the same distribution. Open-set semi-supervised learning (Open-set SSL) considers a more practical scenario, where unlabeled data and test data contain new categories (outliers) not observed in labeled data (inliers). Most previous works focused on outlier detection via binary classifiers, which suffer from insufficient scalability and inability to distinguish different types of uncertainty. In this paper, we propose a novel framework, Adaptive Negative Evidential Deep Learning (ANEDL) to tackle these limitations. Concretely, we first introduce evidential deep learning (EDL) as an outlier detector to quantify different types of uncertainty, and design different uncertainty metrics for self-training and inference. Furthermore, we propose a novel adaptive negative optimization strategy, making EDL more tailored to the unlabeled dataset containing both inliers and outliers. As demonstrated empirically, our proposed method outperforms existing state-of-the-art methods across four datasets.
title Adaptive Negative Evidential Deep Learning for Open-set Semi-supervised Learning
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2303.12091