Deep Label Distribution Learning With Label Ambiguity

DOI10.1109/TIP.2017.2689998MaRDI QIDQ4618849zbMATH OpenWikidataFDO

Authors Bin-Bin Gao, Chao Xing, Chen-Wei Xie, Jianxin Wu, Xin Geng

Publication date 5 February 2019

Published in IEEE Transactions on Image Processing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1611.01731

Mathematics Subject Classification ID

Image processing (compression, reconstruction, etc.) in information and communication theory (94A08)

Abstract: Convolutional Neural Networks (ConvNets) have achieved excellent recognition performance in various visual recognition tasks. A large labeled training set is one of the most important factors for its success. However, it is difficult to collect sufficient training images with precise labels in some domains such as apparent age estimation, head pose estimation, multi-label classification and semantic segmentation. Fortunately, there is ambiguous information among labels, which makes these tasks different from traditional classification. Based on this observation, we convert the label of each image into a discrete label distribution, and learn the label distribution by minimizing a Kullback-Leibler divergence between the predicted and ground-truth label distributions using deep ConvNets. The proposed DLDL (Deep Label Distribution Learning) method effectively utilizes the label ambiguity in both feature learning and classifier learning, which help prevent the network from over-fitting even when the training set is small. Experimental results show that the proposed approach produces significantly better results than state-of-the-art methods for age estimation and head pose estimation. At the same time, it also improves recognition performance for multi-label classification and semantic segmentation tasks.

Cited in

(10)

• Label distribution learning by regularized sample self-representation
• A three-way selective ensemble model for multi-label classification
• Label distribution learning with noisy labels via three-way decisions
• Efficiently learning structured distributions from untrusted batches
• Advances in Intelligent Data Analysis VI
• Classifying vaguely labeled data based on evidential fusion
• Relevance-based label distribution feature selection via convex optimization
• Exploiting asymmetric influence between instances for label enhancement
• Feature selection for multi-labeled data based on label enhancement technique and mutual information
• Learning with confidence: training better classifiers from soft labels

This page was built for publication: Deep Label Distribution Learning With Label Ambiguity

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4618849)