The Emerging Trends of Multi-Label Learning

Liu, Weiwei; Wang, Haobo; Shen, Xiaobo; Tsang, Ivor W.

doi:10.1109/tpami.2021.3119334

Cited by 157 publications

(62 citation statements)

References 151 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multi-label classification has been studied extensively in the last decade (Zhang and Zhou 2014;Liu et al 2021). Most approaches focus on modelling the label correlations to facilitate the learning process, since the output space is exponential in size to the number of class labels.…”

Section: Related Workmentioning

confidence: 99%

“…Multi-label classification deals with the problem where an instance can be associated with multiple labels simultaneously (Zhang and Zhou 2014;Liu et al 2021). As a learning paradigm that handles objects with multiple semantics, researches on multi-label classification have been widely driven by real-world applications, such as multimedia annotation (You et al 2020), text categorization (Tang et al 2020), and bioinformatics analysis (Chen et al 2017), etc.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

End-to-End Probabilistic Label-Specific Feature Learning for Multi-Label Classification

Hang

Zhang

Feng

et al. 2022

AAAI

View full text Add to dashboard Cite

Label-specific features serve as an effective strategy to learn from multi-label data with tailored features accounting for the distinct discriminative properties of each class label. Existing prototype-based label-specific feature transformation approaches work in a three-stage framework, where prototype acquisition, label-specific feature generation and classification model induction are performed independently. Intuitively, this separate framework is suboptimal due to its decoupling nature. In this paper, we make a first attempt towards a unified framework for prototype-based label-specific feature transformation, where the prototypes and the label-specific features are directly optimized for classification. To instantiate it, we propose modelling the prototypes probabilistically by the normalizing flows, which possess adaptive prototypical complexity to fully capture the underlying properties of each class label and allow for scalable stochastic optimization. Then, a label correlation regularized probabilistic latent metric space is constructed via jointly learning the prototypes and the metric-based label-specific features for classification. Comprehensive experiments on 14 benchmark data sets show that our approach outperforms the state-of-the-art counterparts.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

End-to-End Probabilistic Label-Specific Feature Learning for Multi-Label Classification

Hang

Zhang

Feng

et al. 2022

AAAI

View full text Add to dashboard Cite

show abstract

“…Partial annotations. Collecting exhaustive multi-label classification annotations on a large number of classes and images can be intractable, which is why many large-scale datasets resort to partial annotations [31]. For instance, for each image in OpenImages [26] and LVIS [16], only a small fraction of the labels are annotated as positives or negatives.…”

Section: Related Workmentioning

confidence: 99%

Spatial Consistency Loss for Training Multi-Label Classifiers from Single-Label Annotations

Verelst¹,

Rubenstein²,

Eichner³

et al. 2022

Preprint

View full text Add to dashboard Cite

As natural images usually contain multiple objects, multilabel image classification is more applicable "in the wild" than singlelabel classification. However, exhaustively annotating images with every object of interest is costly and time-consuming. We aim to train multilabel classifiers from single-label annotations only. We show that adding a consistency loss, ensuring that the predictions of the network are consistent over consecutive training epochs, is a simple yet effective method to train multi-label classifiers in a weakly supervised setting. We further extend this approach spatially, by ensuring consistency of the spatial feature maps produced over consecutive training epochs, maintaining perclass running-average heatmaps for each training image. We show that this spatial consistency loss further improves the multi-label mAP of the classifiers. In addition, we show that this method overcomes shortcomings of the "crop" data-augmentation by recovering correct supervision signal even when most of the single ground truth object is cropped out of the input image by the data augmentation. We demonstrate gains of the consistency and spatial consistency losses over the binary crossentropy baseline, and over competing methods, on MS-COCO and Pascal VOC. We also demonstrate improved multi-label classification mAP on ImageNet-1K using the ReaL multi-label validation set.

show abstract

“…Multi-label classification is an important and well studied problem [66,62,36] with applications in natural language processing [27,28], audio classification [2,5], information retrieval [46], and computer vision [65,22,15,57,61]. The conventional approach in vision is to train deep convolution neural networks with multiple output predictions -one for each concept/class of interest.…”

Section: Related Workmentioning

confidence: 99%

Multi-Label Learning from Single Positive Labels

Cole¹,

Aodha²,

Lorieul³

et al. 2021

Preprint

View full text Add to dashboard Cite

Predicting all applicable labels for a given image is known as multi-label classification. Compared to the standard multi-class case (where each image has only one label), it is considerably more challenging to annotate training data for multi-label classification. When the number of potential labels is large, human annotators find it difficult to mention all applicable labels for each training image. Furthermore, in some settings detection is intrinsically difficult e.g. finding small object instances in high resolution images. As a result, multi-label training data is often plagued by false negatives. We consider the hardest version of this problem, where annotators provide only one relevant label for each image. As a result, training sets will have only one positive label per image and no confirmed negatives. We explore this special case of learning from missing labels across four different multi-label image classification datasets for both linear classifiers and end-to-end finetuned deep networks. We extend existing multi-label losses to this setting and propose novel variants that constrain the number of expected positive labels during training. Surprisingly, we show that in some cases it is possible to approach the performance of fully labeled classifiers despite training with significantly fewer confirmed labels.

show abstract

The Emerging Trends of Multi-Label Learning

Cited by 157 publications

References 151 publications

End-to-End Probabilistic Label-Specific Feature Learning for Multi-Label Classification

End-to-End Probabilistic Label-Specific Feature Learning for Multi-Label Classification

Spatial Consistency Loss for Training Multi-Label Classifiers from Single-Label Annotations

Multi-Label Learning from Single Positive Labels

Contact Info

Product

Resources

About