Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Generic Multi-label Annotation via Adaptive Graph and Marginalized Augmentation

Generic Multi-label Annotation via Adaptive Graph and Marginalized Augmentation Multi-label learning recovers multiple labels from a single instance. It is a more challenging task compared with single-label manner. Most multi-label learning approaches need large-scale well-labeled samples to achieve high accurate performance. However, it is expensive to build such a dataset. In this work, we propose a generic multi-label learning framework based on Adaptive Graph and Marginalized Augmentation (AGMA) in a semi-supervised scenario. Generally speaking, AGMA makes use of a small amount of labeled data associated with a lot of unlabeled data to boost the learning performance. First, an adaptive similarity graph is learned to effectively capture the intrinsic structure within the data. Second, marginalized augmentation strategy is explored to enhance the model generalization and robustness. Third, a feature-label autoencoder is further deployed to improve inferring efficiency. All the modules are jointly trained to benefit each other. State-of-the-art benchmarks in both traditional and zero-shot multi-label learning scenarios are evaluated. Experiments and ablation studies illustrate the accuracy and efficiency of our AGMA method. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ACM Transactions on Knowledge Discovery from Data (TKDD) Association for Computing Machinery

Generic Multi-label Annotation via Adaptive Graph and Marginalized Augmentation

Download PDF
20 pages

Loading next page...
 
/lp/association-for-computing-machinery/generic-multi-label-annotation-via-adaptive-graph-and-marginalized-KtabFf0g7y

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Association for Computing Machinery
Copyright
Copyright © 2021 Association for Computing Machinery.
ISSN
1556-4681
eISSN
1556-472X
DOI
10.1145/3451884
Publisher site
See Article on Publisher Site

Abstract

Multi-label learning recovers multiple labels from a single instance. It is a more challenging task compared with single-label manner. Most multi-label learning approaches need large-scale well-labeled samples to achieve high accurate performance. However, it is expensive to build such a dataset. In this work, we propose a generic multi-label learning framework based on Adaptive Graph and Marginalized Augmentation (AGMA) in a semi-supervised scenario. Generally speaking, AGMA makes use of a small amount of labeled data associated with a lot of unlabeled data to boost the learning performance. First, an adaptive similarity graph is learned to effectively capture the intrinsic structure within the data. Second, marginalized augmentation strategy is explored to enhance the model generalization and robustness. Third, a feature-label autoencoder is further deployed to improve inferring efficiency. All the modules are jointly trained to benefit each other. State-of-the-art benchmarks in both traditional and zero-shot multi-label learning scenarios are evaluated. Experiments and ablation studies illustrate the accuracy and efficiency of our AGMA method.

Journal

ACM Transactions on Knowledge Discovery from Data (TKDD) – Association for Computing Machinery

Published: Jul 20, 2021

Keywords: Multi-label learning

References

  • Solution of the matrix equation AX+XB=C [F4]
    Bartels, Richard H.; Stewart, G. W.
  • Learning multi-label scene classification
    Boutell, Matthew R.; Luo, Jiebo; Shen, Xipeng; Brown, Christopher M.
  • Distributed optimization and statistical learning via the alternating direction method of multipliers
    Boyd, Stephen; Parikh, Neal; Chu, Eric; Peleato, Borja; Eckstein, Jonathan
  • Convex Optimization
    Boyd, Stephen; Vandenberghe, Lieven
  • Instance annotation for multi-instance multi-label learning
    Briggs, Forrest; Fern, Xiaoli Z.; Raich, Raviv; Lou, Qi
  • New methods for acoustic classification of multiple simultaneous bird species in a noisy environment
    Briggs, F.; Lakshminarayanan, B.; Neal, L.; Fern, X. Z.; Raich, R.; Hadley, S. J. K.; Hadley, A. S.; Betts, M. G.
  • Semi-supervised Learning
    Chapelle, Olivier; Scholkopf, Bernhard; Zien, Alexander
  • Discriminative fields for modeling semantic concepts in video
    Chen, Mingyu; Hauptmann, Alexander
  • Marginalized denoising auto-encoders for nonlinear representations
    Chen, Minmin; Weinberger, Kilian; Sha, Fei; Bengio, Yoshua
  • Fast image tagging
    Chen, Minmin; Zheng, Alice; Weinberger, Kilian
  • Matrix multiplication via arithmetic progressions
    Coppersmith, Don; Winograd, Shmuel
  • ImageNet: A large-scale hierarchical image database
    Deng, Jia; Dong, Wei; Socher, Richard; Li, Lijia; Li, Kai; Li, Feifei
  • Low-rank common subspace for multi-view learning
    Ding, Zhengming; Fu, Yun
  • Semantic autoencoder for zero-shot learning
    Kodirov, Tao Xiang Elyor; Gong, Shagong
  • Collective multi-label classification
    Ghamrawi, Nadia; McCallum, Andrew
  • Discriminative methods for multi-labeled classification
    Godbole, Shantanu; Sarawagi, Sunita
  • TagProp: Discriminative metric learning in nearest neighbor models for image auto-annotation
    Guillaumin, Matthieu; Mensink, Thomas; Verbeek, Jakob; Schmid, Cordelia
  • Semi-supervised multi-label dimensionality reduction
    Guo, Baolin; Hou, Chenping; Nie, Feiping; Yi, Dongyun
  • Leveraging label-specific discriminant mapping features for multi-label learning
    Guo, Yumeng; Chung, Fulai; Li, Guozheng; Wang, Jiancong; Gee, James C.
  • Partially view-aligned clustering
    Huang, Zhenyu; Hu, Peng; Zhou, Joey Tianyi; Lv, Jiancheng; Peng, Xi
  • A shared-subspace learning framework for multi-label classification
    Ji, Shuiwang; Tang, Lei; Yu, Shipeng; Ye, Jieping
  • Robust graph learning from noisy data
    Kang, Z.; Pan, H.; Hoi, S. C. H.; Xu, Z.
  • Exploiting nonlinear relationships for top-N recommender systems
    Kang, Zhao; Peng, Chong; Yang, Ming; Cheng, Qiang
  • Low-rank kernel learning for graph-based clustering
    Kang, Zhao; Wen, Liangjian; Chen, Wenyu; Xu, Zenglin
  • Learning to detect unseen object classes by between-class attribute transfer
    Lampert, Christoph H.; Nickisch, Hannes; Harmeling, Stefan
  • Attribute-based classification for zero-shot visual object categorization
    Lampert, Christoph H.; Nickisch, Hannes; Harmeling, Stefan
  • Learning with marginalized corrupted features and labels together
    Li, Yingming; Yang, Ming; Xu, Zenglin; Zhang, Zhongfei
  • An adaptive graph model for automatic image annotation
    Liu, Jing; Li, Mingjing; Ma, Weiying; Liu, Qingshan; Lu, Hanqing
  • An easy-to-hard learning paradigm for multiple classes and multiple labels
    Liu, Weiwei; Tsang, Ivor W.; Müller, Klaus-Robert
  • Metric learning for multi-output tasks
    Liu, Weiwei; Xu, Donna; Tsang, Ivor; Zhang, Wenjie
  • Optimal lockdown for pandemic control
    Ma, Qianqian; Liu, Yang-Yu; Olshevsky, Alex
  • Adversarial crowdsourcing through robust rank-one matrix completion
    Ma, Qianqian; Olshevsky, Alex
  • Learning with marginalized corrupted features
    Maaten, Laurens; Chen, Minmin; Tyree, Stephen; Weinberger, Kilian
  • Marginalizing corrupted features
    Maaten, Laurens; Chen, Minmin; Tyree, Stephen; Weinberger, Kilian
  • Multi-view clustering and semi-supervised classification with adaptive neighbours
    Nie, Feiping; Cai, Guohao; Li, Xuelong
  • Initialization independent clustering with actively self-training method
    Nie, Feiping; Xu, Dong; Li, Xuelong
  • A general framework for auto-weighted feature selection via global redundancy minimization
    Nie, Feiping; Yang, Sheng; Zhang, Rui; Li, Xuelong
  • Unsupervised feature selection with structured graph optimization
    Nie, Feiping; Zhu, Wei; Li, Xuelong
  • Zero-shot learning with semantic output codes
    Palatucci, Mark; Pomerleau, Dean; Hinton, Geoffrey E.; Mitchell, Tom M.
  • SUN attribute database: Discovering, annotating, and recognizing scene attributes
    Patterson, Genevieve; Hays, James
  • Deep clustering with sample-assignment invariance prior
    Peng, Xi; Zhu, Hongyuan; Feng, Jiashi; Shen, Chunhua; Zhang, Haixian; Zhou, Joey Tianyi
  • Correlative multi-label video annotation
    Qi, Guojun; Hua, Xiansheng; Rui, Yong; Tang, Jinhui; Mei, Tao; Zhang, Hongjiang
  • Multi-label punitive KNN with self-adjusting memory for drifting data streams
    Roseberry, Martha; Krawczyk, Bartosz; Cano, Alberto
  • Very deep convolutional networks for large-scale image recognition
    Simonyan, Karen; Zisserman, Andrew
  • Going deeper with convolutions
    Szegedy, Christian; Liu, Wei; Jia, Yangqing; Sermanet, Pierre; Reed, Scott; Anguelov, Dragomir; Erhan, Dumitru; Vanhoucke, Vincent; Rabinovich, Andrew
  • Multi-label classification of music into emotions
    Trohidis, Konstantinos; Tsoumakas, Grigorios; Kalliris, George; Vlahavas, Ioannis P.
  • The Caltech-UCSD Birds-200-2011 Dataset
    Wah, C.; Branson, S.; Welinder, P.; Perona, P.; Belongie, S.
  • Adaptive graph guided embedding for multi-label annotation
    Wang, Lichen; Ding, Zhengming; Fu, Yun
  • Learning transferable subspace for human motion segmentation
    Wang, Lichen; Ding, Zhengming; Fu, Yun
  • Low-rank transfer human motion segmentation
    Wang, Lichen; Ding, Zhengming; Fu, Yun
  • Generative multi-view human action recognition
    Wang, Lichen; Ding, Zhengming; Tao, Zhiqiang; Liu, Yunyu; Fu, Yun
  • Flexible manifold learning with optimal graph for image and video representation
    Wang, Wei; Yan, Yan; Nie, Feiping; Yan, Shuicheng; Sebe, Nicu
  • Caltech-UCSD Birds 200
    Welinder, Peter; Branson, Steve; Mita, Takeshi; Wah, Catherine; Schroff, Florian; Belongie, Serge; Perona, Pietro
  • ML-MG: Multi-label learning with missing labels using a mixed graph
    Wu, Baoyuan; Lyu, Siwei; Ghanem, Bernard
  • Improving multi-label learning with missing labels by structured semantic correlations
    Yang, Hao; Zhou, Joey Tianyi; Cai, Jianfei
  • Graph-based semi-supervised learning with multiple labels
    Zha, Zhengjun; Mei, Tao; Wang, Jingdong; Wang, Zengfu; Hua, Xiansheng
  • Multilabel relationship learning
    Zhang, Yu; Yeung, Dityan
  • Multilabel dimensionality reduction via dependence maximization
    Zhang, Yin; Zhou, Zhihua
  • Multi-class heterogeneous domain adaptation
    Zhou, Joey Tianyi; Tsang, Ivor W.; Pan, Sinno Jialin; Tan, Mingkui
  • Transfer hashing: From shallow to deep
    Zhou, Joey Tianyi; Zhao, Heng; Peng, Xi; Fang, Meng; Qin, Zheng; Goh, Rick Siow Mong
  • Time-consistent self-supervision for semi-supervised learning
    Zhou, Tianyi; Wang, Shengjie; Bilmes, Jeff
  • Semi-supervised learning literature survey
    Zhu, Xiaojin
  • Semi-supervised learning using Gaussian fields and harmonic functions
    Zhu, Xiaojin; Ghahramani, Zoubin; Lafferty, John D.