Label driven contrastive fusion for multi-view multi-label learning

Abstract

Multi-view multi-label learning (MVML) aims to leverage heterogeneous features and semantic labels for robust model learning. While existing methods often integrate cross-view consensus and view-specific information at the data level, they typically overlook the rich semantics of labels and the inherent view-label correspondences, leading to sub-optimal performance. In this paper, we propose a novel Label-Driven Contrastive Fusion (LDCF) method, which explicitly embeds multi-label semantic correlations into a contrastive fusion scheme with cross-view commonalities exploitation and view-specific individualities extraction. Specifically, LDCF first disentangles consensus and view-specific features through an orthogonal constraint. Then, a label-driven feature selector is designed to construct contrastive sample pairs based on inter-instance label similarities, by which both intra-view and inter-view contrastive learning are applied, pulling semantically similar pairs closer while pushing dissimilar ones apart, thus enhancing the feature discriminability. Finally, LDCF proposes to jointly optimize the orthogonal constraint, the multi-view contrastive learning objective, and the multi-label BCE loss function through a multi-head collaborative classification framework. Extensive experiments on multiple datasets demonstrate the superior performance of the proposed method against state-of-the-art approaches.

Graphical abstract

Keywords

label-driven feature selector / label-driven contrastive fusion / multi-head collaborative classification

References

[1] Zhang C, Yu Z, Hu Q, Zhu P, Liu X, Wang X. Latent semantic aware multi-view multi-label classification. In: Proceedings of the 32nd AAAI Conference on Artificial Intelligence. 2018

[2] Zhang W, Li M, Chen Y, Wang H . Cross-view prototype contrastive learning for multi-view multi-label classification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024, 46( 3): 1521–1535

[3] Peng G W, Dornaika , F D, Charafeddine J N . Multi-view learning with graph convolution networks adopting diverse graphs and genuine deep feature fusion. IEEE Transactions on Neural Networks and Learning Systems, 2025, 58( 9): 1573–7462

[4] Luo Y, Liu T, Tao D, Xu C . Multiview matrix completion for multilabel image classification. IEEE Transactions on Image Processing, 2015, 24( 8): 2355–2368

[5] Tan Q, Yu G, Wang J, Domeniconi C, Zhang X . Individuality-and commonality-based multiview multilabel learning. IEEE Transactions on Cybernetics, 2021, 51( 3): 1716–1727

[6] Lyu G, Deng X, Wu Y, Feng S. Beyond shared subspace: a view-specific fusion for multi-view multi-label learning. In: Proceedings of the 36th AAAI Conference on Artificial Intelligence. 2022, 7647–7654

[7] Xu K, Zhang R, Li S . Adversarial fusion for balancing shared and view-specific information in multi-view multi-label learning. IEEE Transactions on Image Processing, 2024, 33: 1124–1137

[8] Lyu G, Yang Z, Deng X, Feng S . L-VSM: label-driven view-specific fusion for multiview multilabel classification. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36( 4): 6569–6583

[9] Ma R, Li S, Zhang B, Fang L, Li Z . Flexible and generalized real photograph denoising exploiting dual meta attention. IEEE Transactions on Cybernetics, 2023, 53( 10): 6395–6407

[10] Shu X, Xu B, Zhang L, Tang J . Multi-granularity anchor-contrastive representation learning for semi-supervised skeleton-based action recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45( 6): 7559–7576

[11] Zhang M L, Wu L . Lift: Multi-label learning with label-specific features. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37( 1): 107–120

[12] Zhang M L, Li Y K, Liu X Y, Geng X . Binary relevance for multi-label learning: an overview. Frontiers of Computer Science, 2018, 12( 2): 191–202

[13] Sun L, Feng S, Lyu G, Zhang H, Dai G . Partial multi-label learning with noisy side information. Knowledge and Information Systems, 2021, 63( 2): 541–564

[14] Li Y, Zhang H, Wu C . Graph-guided pairwise label ranking for efficient multi-label learning. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35( 3): 1124–1137

[15] Lyu G, Feng S, Jin Y, Wang T, Lang C, Li Y . Prior knowledge regularized self-representation model for partial multilabel learning. IEEE Transactions on Cybernetics, 2023, 53( 3): 1618–1628

[16] Luo T, Li Q, Jiang M, Hou C . Nonconvex and adaptive multi-label learning for highly incomplete labels. Knowledge-Based Systems, 2026, 331: 114784

[17] Chen Z M, Wei X S, Wang P, Guo Y. Multi-label image recognition with graph convolutional networks. In: Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019, 5172–5181

[18] Huang J, Li G, Huang Q, Wu X. Learning label specific features for multi-label classification. In: Proceedings of 2015 IEEE International Conference on Data Mining. 2015, 181–190

[19] Gao W, Hao P, Wu Y, Zhang P . A unified low-order information-theoretic feature selection framework for multi-label learning. Pattern Recognition, 2023, 134: 109111

[20] Zhang X, Luo T, Liu Y, Hou C. Imbalanced multi-instance multi-label learning via coding ensemble and adaptive thresholds. In: Proceedings of the 32nd ACM International Conference on Multimedia. 2024, 5413–5422

[21] Liu I J, Lin C L, Yang F E, Wang Y C F. Language-guided transformer for federated multi-label classification. In: Proceedings of the 38th AAAI Conference on Artificial Intelligence. 2024, 13882–13890

[22] Zhang P, Wu M. Multi-label supervised contrastive learning. In: Proceedings of the 38th AAAI Conference on Artificial Intelligence. 2024, 16786–16793

[23] Zhang S, Xu R, Xiong C, Ramaiah C. Use all the labels: a hierarchical multi-label contrastive learning framework. In: Proceedings of 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022, 16639–16648

[24] Jiang X, He B, Zhou P Y, Chen X, Guo J, Xu J, Liao Y. A unified framework to bridge complete and incomplete deep multi-view clustering under non-IID missing patterns. In: Proceedings of 2025 IEEE International Conference on Computer Vision. 2025, 594–603

[25] Liu M, Jiang W, Xu J, others . Two-level semantic consistency hashing for scalable multi-label image retrieval. Frontiers of Computer Science, 2025, 19( 3): 193318

[26] Li X, Chen S . A concise yet effective model for non-aligned incomplete multi-view and missing multi-label learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 44( 10): 5918–5932

[27] Yu Z, Dong Z, Yu C, Yang K, Fan Z, Chen C L P . A review on multi-view learning. Frontiers of Computer Science, 2025, 19( 7): 197334

[28] Wang Q, Tao Z, Gao Q, Jiao L . Multi-view subspace clustering via structured multi-pathway network. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35( 5): 7244–7250

[29] Zong L, Zhang X, Zhao L, Yu H, Zhao Q . Multi-view clustering via multi-manifold regularized non-negative matrix factorization. Neural Networks, 2017, 88: 74–89

[30] Li L, He H . Bipartite graph based multi-view clustering. IEEE Transactions on Knowledge and Data Engineering, 2022, 34( 7): 3111–3125

[31] He H, Xu J, Wen G, Ren Y, Zhao N, Zhu X. Graph embedded contrastive learning for multi-view clustering. In: Proceedings of the 34th International Joint Conference on Artificial Intelligence. 2025, 5336–5344

[32] Liu J Y, Liu X W, Yang Y X, Guo X F, Kloft M, He L Z. Multiview Subspace Clustering via Co-Training Robust Data Representation. IEEE Transactions on Neural Networks and Learning Systems. 2025, 33(10): 5177−5189.

[33] Yan X, Wang S, Chen H, Zhu H . Multi-view learning with enhanced multi-weight vector projection support vector machine. Neural Networks, 2025, 185: 107180

[34] Luo Y, Zhang L, Liu W, Tao D . Multi-view learning with correntropyinduced loss. Frontiers of Computer Science, 2019, 13( 5): 1062–1076

[35] Lin Y, Wang Y, Lyu G, Deng Y, Cai H, Lin H, Wang H, Yang Z. Enhance multi-view classification through multi-scale alignment and expanded boundary. In: Proceedings of the 13th International Conference on Learning Representations. 2025

[36] Huang Z, Zhou J T, Zhu H, Zhang C, Lv J, Peng X . Deep spectral representation learning from multi-view data. IEEE Transactions on Image Processing, 2021, 30: 5352–5362

[37] Liu W, Chen Y, Yue X. Enhancing multi-view classification reliability with adaptive rejection. In: Proceedings of the 39th AAAI Conference on Artificial Intelligence. 2025, 18969–18977

[38] Xu J, Zhao N, Niu G, Sugiyama M, Zhu X. Robust multi-view learning via representation fusion of sample-level attention and alignment of simulated perturbation. In: Proceedings of IEEE/CVF International Conference on Computer Vision. 2025, 4233–4241

[39] Zhong Q, Lyu G, Yang Z . Align while fusion: a generalized nonaligned multiview multilabel classification method. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36( 4): 7627–7636

[40] Li Q, Luo T, Jiang M, Jiang Z, Hou C, Li F. Semi-supervised multi-view multi-label learning with view-specific transformer and enhanced pseudo-label. In: Proceedings of the 39th AAAI Conference on Artificial Intelligence. 2025, 18430–18438

[41] Liu M, Luo Y, Tao D, Xu C, Wen Y G. Low-rank multi-view learning in matrix completion for multi-label image classification. In: Proceedings of the 29th AAAI Conference on Artificial Intelligence. 2015, 2778–2784

[42] Hao P, Liu K, Gao W. Uncertainty-aware global-view reconstruction for multi-view multi-label feature selection. In: Proceedings of the 39th AAAI Conference on Artificial Intelligence. 2025, 17068–17076

[43] Wu X, Chen Q G, Hu Y, Wang D, Chang X, Wang X, Zhang M L. Multi-view multi-label learning with view-specific information extraction. In: Proceedings of the 28th International Joint Conference on Artificial Intelligence. 2019, 3884–3890

[44] Zhao D, Gao Q, Lu Y, Sun D, Cheng Y . Consistency and diversity neural network multi-view multi-label learning. Knowledge-Based Systems, 2021, 218: 106841

[45] Hao P, Liu K, Gao W. Double-layer hybrid-label identification feature selection for multi-view multi-label learning. In: Proceedings of the 38th AAAI Conference on Artificial Intelligence. 2024, 12295–12303

[46] Wen J, Liu Y D, Tang Z Y, He Y T, Chen Y L, Li M, Liu C L. In: proceedings of International Conference on Machine Learning. 2025, 66467−66480.

[47] Kingma D P, Ba J. Adam: a method for stochastic optimization. In: Proceedings of the 3rd International Conference on Learning Representations. 2015

[48] Sun S, Zong D . LCBM: a multi-view probabilistic model for multi-label classification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43( 8): 2682–2696

[49] Szymański P, Kajdanowicz T, Kersting K . How is a data-driven approach better than random choice in label space division for multi-label classification?. Entropy, 2016, 18( 8): 282

[50] Wu J H, Wu X, Chen Q G, Hu Y, Zhang M L. Feature-induced manifold disambiguation for multi-view partial multi-label learning. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2020, 557–565

[51] Lu X, Feng S, Lyu G, Jin Y, Lang C . Distance-preserving embedding adaptive bipartite graph multi-view learning with application to multi-label classification. ACM Transactions on Knowledge Discovery from Data, 2023, 17( 2): 19

[52] Wen J, Liu C, Deng S, Liu Y, Fei L, Yan K, Xu Y . Deep double incomplete multi-view multi-label learning with incomplete labels and missing views. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35( 8): 11396–11408

[53] Lyu G, Kang W, Wang H, Li Z, Yang Z, Feng S. Common-individual semantic fusion for multi-view multi-label learning. In: Proceedings of the 33rd International Joint Conference on Artificial Intelligence. 2024, 4715–4723

[54] Wei H, Deng Y, Hai Q, Lin Y, Yang Z, Lyu G. Multi-view multi-label classification via view-label matching selection. In: Proceedings of the 39th AAAI Conference on Artificial Intelligence. 2025, 21456–21464

[55] Demšar J . Statistical comparisons of classifiers over multiple data sets. Journal of Machine Learning Research, 2006, 7( 1): 1–30

RIGHTS & PERMISSIONS

Higher Education Press