Multi-label learning with missing and completely unobserved labels

Huang, Jun; Xu, Linchuan; Qian, Kun; Wang, Jing; Yamanishi, Kenji

doi:10.1007/s10618-021-00743-x

“…Label Completion. There is a rich body of work in the literature that focuses on learning from partial/missing labels in a multi-label learning setting where each image is labelled for multiple categories with some missing labels [5,6,8,14,20,34]. Common strategies to address this can be divided into two categories: 1) graph based methods [24,54] that exploit similarity between samples to predict missing labels, and 2) low rank matrix completion which extracts label correlations [6,14,56,59] to complete missing labels.…”

Section: Related Workmentioning

confidence: 99%

Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation

Goel

¹

,

Fernando

²

,

Keller

³

et al. 2022

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

View full text Add to dashboard Cite

Scene graph generation (SGG) aims to capture a wide variety of interactions between pairs of objects, which is essential for full scene understanding. Existing SGG methods trained on the entire set of relations fail to acquire complex reasoning about visual and textual correlations due to various biases in training data. Learning on trivial relations that indicate generic spatial configuration like 'on' instead of informative relations such as 'parked on' does not enforce this complex reasoning, harming generalization. To address this problem, we propose a novel framework for SGG training that exploits relation labels based on their informativeness. Our model-agnostic training procedure imputes missing informative relations for less informative samples in the training data and trains a SGG model on the imputed labels along with existing annotations. We show that this approach can successfully be used in conjunction with state-of-the-art SGG methods and improves their performance significantly in multiple metrics on the standard Visual Genome benchmark. Furthermore, we obtain considerable improvements for unseen triplets in a more challenging zero-shot setting.

show abstract

“…The datasets including enron, medical, Corel5k and Corel16k001 are selected from Mulan [24], which is an open Java library for multi-label learning 1 . The selected datasets can evaluate the proposed method in different cases including text and image.…”

Section: A Datasetsmentioning

confidence: 99%

“…R ECENT years have witnessed many approaches to solve the problem of that one object may associate with a set of labels, which is also commonly framed as the multi-label classification problem [1]. Different from binary class and multi-class classification in single-label problem, the intrinsic multi-label nature of most real datasets could represent the world more exactly [2]- [4].…”

Section: Introductionmentioning

confidence: 99%

Representation Learning With Dual Autoencoder for Multi-Label Classification

Zhu

¹

,

Yang

²

,

Li

³

et al. 2021

View full text Add to dashboard Cite

Multi-label classification aims to deal with the problem that an object may be associated with one or more labels, which is a more difficult task due to the complex nature of multi-label data. The crucial problem of multi-label classification is the more robust and higher-level feature representation learning, which can reduce non-helpful feature attributes from the input space prior to training. In recent years, deep learning methods based on autoencoders have achieved excellent performance in multi-label classification for the advantages of powerful representations learning ability and fast convergence speed. However, most existing autoencoder-based methods only rely on the single autoencoder model, which pose challenges for multi-label feature representations learning and fail to measure similarities between data spaces. To address this problem, in this paper, we propose a novel representation learning method with dual autoencoder for multi-label classification. Compared to the existing autoencoder-based methods, our proposed method can capture different characteristics and more abstract features from data by the serially connection of two different types of autoencoders. More specifically, firstly, the algorithm of Reconstruction Independent Component Analysis (RICA) in sparse autoencoder is trained on patches on all training and test dataset for robust global feature representations learning. Secondly, with the output of RICA, stacked autoencoder with manifold regularization (SAMR) is introduced to ameliorate the quality of multi-label features learning. Comprehensive experiments on several real-world data sets demonstrate the effectiveness of our proposed approach compared with several competing state-of-the-art methods.

show abstract

“…Label Completion. There is a rich body of work in the literature that focuses on learning from partial/missing labels in a multi-label learning setting where each image is labelled for multiple categories with some missing labels [5,6,8,14,20,34]. Common strategies to address this can be divided into two categories: 1) graph based methods [24,54] that exploit similarity between samples to predict missing labels, and 2) low rank matrix completion which extracts label correlations [6,14,56,59] to complete missing labels.…”

Section: Related Workmentioning

confidence: 99%

Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation

Goel¹,

Fernando²,

Keller³

et al. 2021

Preprint

0

View full text Add to dashboard Cite

Scene graph generation (SGG) aims to capture a wide variety of interactions between pairs of objects, which is essential for full scene understanding. Existing SGG methods trained on the entire set of relations fail to acquire complex reasoning about visual and textual correlations due to various biases in training data. Learning on trivial relations that indicate generic spatial configuration like 'on' instead of informative relations such as 'parked on' does not enforce this complex reasoning, harming generalization. To address this problem, we propose a novel framework for SGG training that exploits relation labels based on their informativeness. Our model-agnostic training procedure imputes missing informative relations for less informative samples in the training data and trains a SGG model on the imputed labels along with existing annotations. We show that this approach can successfully be used in conjunction with state-of-the-art SGG methods and improves their performance significantly in multiple metrics on the standard Visual Genome benchmark. Furthermore, we obtain considerable improvements for unseen triplets in a more challenging zero-shot setting.

show abstract

Multi-label learning with missing and completely unobserved labels

Cited by 22 publications

References 36 publications

Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation

Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation

Representation Learning With Dual Autoencoder for Multi-Label Classification

Not All Relations are Equal: Mining Informative Labels for Scene Graph Generation

Contact Info

Product

Resources

About