Transductive Multi-View Zero-Shot Learning
Open accessThe DOI matches its Crossref record; 72 of the 74 references it cites were found in a registry or an index.
Abstract
Most existing zero-shot learning approaches exploit transfer learning via an intermediate semantic representation shared between an annotated auxiliary dataset and a target dataset with different classes and no annotation. A projection from a low-level feature space to the semantic representation space is learned from the auxiliary dataset and applied without adaptation to the target dataset. In this paper we identify two inherent limitations with these approaches. First, due to having disjoint and potentially unrelated classes, the projection functions learned from the auxiliary dataset/domain are biased when applied directly to the target dataset/domain. We call this problem the projection domain shift problem and propose a novel framework, transductive multi-view embedding, to solve it. The second limitation is the prototype sparsity problem which refers to the fact that for each target class, only a single prototype is available for zero-shot learning given a semantic representation. To overcome this problem, a novel heterogeneous multi-view hypergraph label propagation method is formulated for zero-shot learning in the transductive embedding space. It effectively exploits the complementary information offered by different semantic representations and takes advantage of the manifold structures of multiple representation spaces in a coherent manner. We demonstrate through extensive experiments that the proposed approach (1) rectifies the projection shift between the auxiliary and target domains, (2) exploits the complementarity of multiple semantic representations, (3) significantly outperforms existing methods for both zero-shot and N-shot recognition on three image and video benchmark datasets, and (4) enables novel cross-view annotation tasks.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“What is interesting to see here is that Overfeat > Decaf since both are based on the same Convolutional Neural Network (CNN) model of [32].”
BackgroundCollective dynamics of ‘small-world’ networks
Watts, D. J., & Strogatz, S. H. (1998) · Nature · doi:10.1038/30918 · Cited by 43,060
LIBSVM: A library for support vector machines
Chang, C.-C., & Lin, C.-J. (2011) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/1961189.1961199 · Cited by 40,918 · cited 1×
“However, this does not effectively exploit the multi-view complementarity, and suffers from labelled data (prototype) sparsity.”
Background(2008) · Journal of Machine Learning Research · Cited by 33,051
Pan, S. J., & Yang, Q. (2010) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2009.191 · Cited by 23,387
Learning Deep Architectures for AI
Bengio, Y. (2009) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000006 · Cited by 9,239 · cited 1×
“The advantages of heterogeneous hypergraphs We argue that the pairwise similarity of heterogeneous hypergraph is a distributed representation [3].”
BackgroundRecognition-by-components: A theory of human image understanding.
Biederman, I. (1987) · Psychological Review · doi:10.1037/0033-295x.94.2.115 · Cited by 5,621 · cited 1×
“Humans can distinguish 30,000 basic object classes [3] and many more subordinate ones (e.g. breeds of dogs).”
BackgroundLearning Deep Architectures for AI
Bengio, Y. (2009) · now publishers, Inc. eBooks · doi:10.1561/9781601982957 · Cited by 4,941
Semi-Supervised Learning Literature Survey
Zhu, X. (2005) · Minds at UW (University of Wisconsin) · Cited by 4,505
DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition
Donahue, J., Jia, Y., Vinyals, O., Hoffman, J., Zhang, N., Tzeng, E., & Darrell, T. (2013) · arXiv · doi:10.48550/arxiv.1310.1531 · Cited by 4,019 · cited 1×
“In addition, the Decaf (D) [10] feature is also considered for AwA.”
Background
Research works printed but not linked: 5
- [39] E. Rosch. Classification of real-world objects: Origins and representations in cognition. Thinking: Readings in Cognitive Science , 1977. -
- [32] T. Mikolov, K. Chen, G. Corrado, and J. Dean. Efficient estimation of word representation in vector space. In Proceedings of Workshop at ICLR , 2013. -
- [33] M. Palatucci, G. Hinton, D. Pomerleau, and T. M. Mitchell. Zeroshot learning with semantic output codes. In NIPS , 2009. -
- [34] D. Parikh and K. Grauman. Relative attributes. In ICCV , 2011. -
- [21] Y. Huang, Q. Liu, and D. Metaxas. Video object segmentation by hypergraph cut. In CVPR , 2009. -
Cited by
Items 301–400 of 512
Visual Space Optimization for Zero-shot LearningPreprint
Wang, X., Pang, S., Zhu, J., Li, Z., Tian, Z., & Li, Y. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.00330 · Cited by 8
“Furthermore, a related scenario is the transductive zeroshot learning [34], [35], [36], [23] , which assumes that the unlabeled samples from unseen classes are available during training.”
BackgroundChen, X., Zhu, G., & Ji, C. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3313769 · Cited by 7
Cycle-Consistent Generating Network Based on Semantic Distance for Zero-Shot Fault Diagnosis
Liao, W., Wu, L., Xu, S., & Fujimura, S. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3545713 · Cited by 7
Content-Attribute Disentanglement for Generalized Zero-Shot Learning
An, Y., Kim, S., Liang, Y., Zimmermann, R., Kim, D., & Kim, J. (2022) · IEEE Access · doi:10.1109/access.2022.3178800 · Cited by 7
Enhancing Domain-Invariant Parts for Generalized Zero-Shot Learning
Zhang, Y., & Feng, S. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3611764 · Cited by 7
Multidomain Features Fusion for Zero-Shot Learning
Liu, Z., Zeng, Z., & Lian, C. (2018) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2018.2868061 · Cited by 7
“The models which establish visual-semantic projection function only through seen class samples, inevitably encounter the projection domain shift problem [24].”
BackgroundMethodJoint Projection and Subspace Learning for Zero-Shot Recognition
Liu, G., Guan, J., Zhang, M., Zhang, J., Wang, Z., & Lu, Z. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00214 · Cited by 10
“The largest challenge in ZSL is the projection domain shift [6].”
BackgroundHolistically Associated Transductive Zero-Shot Learning
Xu, Y., Xu, X., Han, G., & He, S. (2021) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2021.3049274 · Cited by 7
“Cosine similarity is widely used in many ZSL literatures [3], [13], [25], [26], [50], [67] for its good ability in measuring the similarity between two semantic representations.”
MethodInference Fusion with Associative Semantics for Unseen Object Detection
Li, Y., Li, P., Cui, H. F., & Wang, D. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i3.16295 · Cited by 7
Sentinel-Guided Zero-Shot Learning: A Collaborative Paradigm Without Real Data Exposure
Wan, F., Miao, X., Duan, H., Deng, J., Gao, R., & Long, Y. (2024) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2024.3384756 · Cited by 11
“Depending on the utilization of unseen data during training, ZSL methods can be categorized into inductive [44], [45] and transductive settings [46], [47].”
BackgroundZero-Shot Learning with Superclasses
Huo, Y., Ding, M., Zhao, A., Hu, J. F., Wen, J.-R., & Lu, Z. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-04182-3_40 · Cited by 8
Zero-Shot Facial Expression Recognition with Multi-label Label Propagation
Lu, Z., Zeng, J., Shan, S., & Chen, X. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-20893-6_2 · Cited by 10
Learning the Compositional Spaces for Generalized Zero-shot LearningPreprint
Dong, H., Fu, Y., Hwang, S. J., Sigal, L., & Xue, X. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1810.07368 · Cited by 7
“To enable the knowledge transfer, semantic representations, auxiliary information such as visual attributes [28] and word embeddings [18] are used to relate target classes with source classes.”
MethodSelf-citationSEMANTIC BORROWING FOR GENERALIZED ZERO-SHOT LEARNINGPreprint
Xiao-wei, C. H. E. N. (2021) · doi:10.36227/techrxiv.13683919 · Cited by 7
Zero-Shot Object Detection by Semantics-Aware DETR with Adaptive Contrastive Loss
Liu, H., Zhang, L., Guan, J., & Zhou, S. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3612523 · Cited by 6
Learning object-centric complementary features for zero-shot learning
Liu, J., Song, K., He, Y., Dong, H., Yan, Y., & Meng, Q. (2020) · Signal Processing Image Communication · doi:10.1016/j.image.2020.115974 · Cited by 6
Adversarial strategy for transductive zero-shot learning
Liu, Y., Du, B., & Ni, F. (2021) · Information Sciences · doi:10.1016/j.ins.2021.06.085 · Cited by 6
Confusion-Based Metric Learning for Regularizing Zero-Shot Image Retrieval and Clustering
Chen, B., Deng, W., Wang, B., & Zhang, L. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3185668 · Cited by 6
CoHOZ: Contrastive Multimodal Prompt Tuning for Hierarchical Open-set Zero-shot Recognition
Liao, N., Liu, Y., Li, X., Lei, C., Wang, G., Hua, X., & Yan, J. (2022) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3503161.3548021 · Cited by 9
“In contrast, transductive ZSL methods usually utilize an auxiliary set of class attributes describing the seen classes in the training phase and the unseen classes in the testing phase to transfer the learned knowledge from seen classes to unseen classes [11, 33, 51, 71].”
BackgroundUnbiased feature generating for generalized zero-shot learning
Niu, C., Shang, J., Huang, J., Yang, J., Song, Y., Zhou, Z., & Zhou, G. (2022) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2022.103657 · Cited by 6
A generative adversarial network with “zero-shot” learning for positron image denoising
Zhu, M., Zhao, M., Yao, M., & Guo, R. (2023) · Scientific Reports · doi:10.1038/s41598-023-28094-1 · Cited by 6
“Improved the original model and trained a labeling framework for model embedding directly, which realized the prediction of data categories 27 .”
MethodConsistency-guided pseudo labeling for transductive zero-shot learning
Yang, H., Wang, N., Wang, Z., Wang, L., & Li, H. (2024) · Information Sciences · doi:10.1016/j.ins.2024.120572 · Cited by 9
Meta-DZSL: a meta-dictionary learning based approach to zero-shot recognition
Singh, U. P., Singh, K. P., & Thakur, M. (2022) · Applied Intelligence · doi:10.1007/s10489-022-03257-1 · Cited by 6
“Despite this, many zero-shot learners in inductive settings can be found in the literature [27, 34–37].”
BackgroundDifferential Refinement Network for Zero-Shot Learning
Tian, Y., Zhang, Y., Huang, Y., Xu, W., & Ding, Z. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3201883 · Cited by 6
“Fu et al. [8] first discussed the projection domain shift problem and proposed a transductive multiview embedding to solve it.”
BackgroundA Robust 3-D Reconstruction Approach Based on Correspondence Retrieval Using Deep Image Prior
Li, M., Sun, C., Fu, L., & Wang, P. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3267526 · Cited by 7
“The matching fails frequently as networks optimized using specific dataset can hardly cope with tested images with feature distributions different from the training set [23], [24], [25], [26].”
BackgroundTransfer Feature Generating Networks With Semantic Classes Structure for Zero-Shot Learning
Lin, G., Chen, W., Liao, K., Kang, X., & Fan, C. (2019) · IEEE Access · doi:10.1109/access.2019.2958052 · Cited by 6
Self-Training Ensemble Networks for Zero-Shot Image Recognition.Preprint
Ye, M., & Guo, Y. (2018) · arXiv (Cornell University) · Cited by 6
Dong, H., Fu, Y., Sigal, L., Hwang, S. J., Jiang, Y., & Xue, X. (2018) · arXiv (Cornell University) · Cited by 6
Self-citationCompositional Fine-Grained Low-Shot LearningPreprint
Huynh, D., & Elhamifar, E. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.10438 · Cited by 7
“Moreover, the predictions of these methods are often biased towards seen classes due to the lack of training samples for novel classes [63], [64].”
BackgroundA Cross-Modal Alignment for Zero-Shot Image Classification
Wu, L., Wu, C., Guo, H. W., & Zhao, Z. (2023) · IEEE Access · doi:10.1109/access.2023.3237966 · Cited by 6
“It is also hard to distinguish the feature with same text attribute but different classes [5].”
BackgroundPrototype Calibration with Synthesized Samples for Zero-Shot Chinese Character Recognition
Ao, X., Li, X.-H., Zhang, X.-Y., & Liu, C. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446790 · Cited by 5
“However, all above-mentioned methods suffer from the problem of domain shift [9] which is caused by the fact that the alignment between samples and auxiliary semantic data on seen categories is not always consistent with that on unseen categories.”
BackgroundMulti-Label Zero-Shot Learning With Adversarial and Variational Techniques
Gull, M., & Arif, O. (2024) · IEEE Access · doi:10.1109/access.2024.3425547 · Cited by 5
“Single-label GZSL has garnered significant attention in recent years [24], [25], [26], [27], [28], [29], [30], [31].”
BackgroundZero-Shot Learning Via Recurrent Knowledge Transfer
Zhao, B., Sun, X., Hong, X., Yao, Y., & Wang, Y. (2019) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2019.00144 · Cited by 6
“Similar as many existing works [13, 43, 52], we use official models (VGG, GoogLeNet, and ResNet) pre-trained on ImageNet to extract features.”
BackgroundMethodInfluentialPairnorm based Graphical Convolution Network for zero-shot multi-label classification
Chauhan, V., & Tiwari, A. (2022) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2022.105012 · Cited by 5
Semantic-visual shared knowledge graph for zero-shot learning
Yu, B., Xie, C., Tang, P., & Li, B. (2023) · PeerJ Computer Science · doi:10.7717/peerj-cs.1260 · Cited by 6
“This leads a serious ‘‘domain-shift’’ problem (Min et al., 2020; Fu et al., 2015; Zhu et al., 2018;Wan et al., 2019; Ni, Zhang & Xie, 2019) during projection between semantic space and visual space.”
BackgroundManifold embedding for zero-shot recognition
Ji, Z., Yu, X., Yu, Y., & He, Y. (2018) · Cognitive Systems Research · doi:10.1016/j.cogsys.2018.12.003 · Cited by 5
“By imitating human’s ability to recognize unseen categories with semantic descriptions, Zero-Shot Recogintion(ZSR) [4, 5, 6, 7, 8] is attracting a lot attention in recent years.”
BackgroundBeyond Attributes: High-Order Attribute Features for Zero-Shot Learning
Jin, X., Xie, G.-S., Huang, K., Miao, J., & Wang, Q. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00357 · Cited by 5
Class knowledge overlay to visual feature learning for zero-shot image classification
Xie, C., Zeng, T., Xiang, H., Li, K., Yang, Y., & Liu, Q. (2021) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2021.103206 · Cited by 5
Prototypical Network for Radar Image Recognition with Few Samples
Liu, D., Gao, X., & Shen, Q. (2020) · Journal of Physics Conference Series · doi:10.1088/1742-6596/1634/1/012116 · Cited by 5
“This phenomenon is called domain shift in ZSL[8].”
BackgroundA Novel Approach of Unknown Network Attack Detection Based on Zero-Shot Learning
Wang, H., Wang, Y., & Guo, Y. (2021) · IEEE International Conference on Data Science and Computer Application (ICDSCA) · doi:10.1109/icdsca53499.2021.9650182 · Cited by 5
SDM-Net: A Simple and Effective Model for Generalized Zero-Shot LearningPreprint
Daghaghi, S., Medini, T., & Shrivastava, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.04790 · Cited by 6
Multi-Knowledge Fusion for New Feature Generation in Generalized Zero-Shot LearningPreprint
Xiang, H., Xie, C., Zeng, T., & Yang, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.11566 · Cited by 6
“ZSL approaches [27], [28], [29], [30] have been proposed.”
Yang, C., Jin, Q., Du, F., Guo, J., & Zhou, Y. (2025) · Complex & Intelligent Systems · doi:10.1007/s40747-024-01717-4 · Cited by 4
“Domain shift [11] has been a persistent problem in the field of ZSL.”
Zero-shot learning via self-organizing maps
Ismailoglu, F. (2023) · Neural Computing and Applications · doi:10.1007/s00521-023-08299-1 · Cited by 4
“To overcome this problem, the obvious solution is to incorporate the unlabelled test data into the learning process, which is achieved by a tiny fraction of ZSL methods [11].”
BackgroundMethodStructure Fusion and Propagation for Zero-Shot Learning
Lin, G., Chen, Y., & Zhao, F. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-03338-5_39 · Cited by 4
“These semantic information can not only be regarded as a middle bridge for associating visual images and class labels [14] [15] [16] [17] [18] [19] [20], but also be transformed into new representations corresponding to the more suitable relation between unseen and seen classes by Canonical Correlation Analysis (CCA)[…”
BackgroundMethodConcept Embedding through Canonical Forms: A Case Study on Zero-Shot ASL Recognition
Kamzin, A., Amperayani, V. N. S. A., Sukhapalli, P., Banerjee, A., & Gupta, S. K. S. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412207 · Cited by 4
“Existing solutions to this problem, typically increase the complexity of zero-shot learning with only incremental improvement in accuracy [14].”
BackgroundA zero-shot learning boosting framework via concept-constrained clustering
Yue, Q., Cui, J., Bai, L., Liang, J., & Liang, J. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.109937 · Cited by 4
Open-vocabulary multi-label classification with visual and textual features fusion
Liu, T., Chen, Y., Chen, G., & Li, W. (2024) · The Visual Computer · doi:10.1007/s00371-024-03769-6 · Cited by 4
Li, Y., & Zhang, Y. (2025) · Advances in Civil Engineering · doi:10.1155/adce/9363110 · Cited by 4
“…data spatial distribution relationships., and then further search for the spatial distribution of the data, and in ImageNet [11]; Fu et al. [12] proposed a trans-ductive multiview embedding framework to solve the problem of domain drift, and used heterogeneous multiview label propagation…”
Transductive zero-shot learning via knowledge graph and graph convolutional networks
Li, Q., Sun, X., & Dong, J. (2025) · Scientific Reports · doi:10.1038/s41598-025-13612-0 · Cited by 4
Emerging Topics in Learning from Noisy and Missing Data
Alameda-Pineda, X., Hospedales, T. M., Ricci, E., Sebe, N., & Wang, X. (2016) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/2964284.2986910 · Cited by 4
“Special emphasis will be given to methods based on deep architectures for unsupervised domain adaptation [6, 5], low-rank modeling for learning in transductive settings and zero-shot learning [3].”
Self-citationA Novel Group-Aware Pruning Method for Few-shot Learning
Zheng, Y.-D., Ma, Y., Liu, R.-Z., & Lü, T. (2019) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2019.8852221 · Cited by 4
Dual discriminative auto-encoder network for zero shot image recognition
Bai, H., Zhang, H., & Wang, Q. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-201920 · Cited by 6
JSE: Joint Semantic Encoder for zero-shot gesture learning
Madapana, N., & Wachs, J. P. (2021) · Pattern Analysis and Applications · doi:10.1007/s10044-021-00992-y · Cited by 5
NucNormZSL: nuclear norm-based domain adaptation in zero-shot learning
Singh, U. P., Singh, K. P., & Thakur, M. (2021) · Neural Computing and Applications · doi:10.1007/s00521-021-06461-1 · Cited by 4
Trading-off Information Modalities in Zero-shot Classification
Sanchez, J., & Molina, M. (2022) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv51458.2022.00174 · Cited by 4
Transferability of features for neural networks links to adversarial attacks and defences
Kotyan, S., Matsuki, M., & Vargas, D. V. (2022) · PLoS ONE · doi:10.1371/journal.pone.0266060 · Cited by 4
Prompt-guided Scene Generation for 3D Zero-Shot Learning
Nasiri, M., Cheraghian, A., Chowdhury, T. F., Ahmadi, S., Saberi, M., & Rahman, S. (2022) · International Conference on Digital Image Computing: Techniques and Applications (DICTA) · doi:10.1109/dicta56598.2022.10034623 · Cited by 4
Privacy-Enhanced Zero-Shot Learning via Data-Free Knowledge Transfer
Gao, R., Wan, F., Organisciak, D., Pu, J., Duan, H., Zhang, P., Hou, X., & Long, Y. (2023) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme55011.2023.00081 · Cited by 9
Application of CLIP for efficient zero-shot learning
Yang, H., Wang, N., Li, H., Wang, L., & Wang, Z. (2024) · Multimedia Systems · doi:10.1007/s00530-024-01414-9 · Cited by 4
“In the semantic space, the semantic embeddings of unseen class images deviate from their corresponding class prototypes [3].”
Robust zero-shot learning with source attributes noise
Yu, J., Wu, S., Wang, L., & Jing, X. (2016) · IEEE International Conference on Progress in Informatics and Computing · doi:10.1109/pic.2016.7949495 · Cited by 4
“by jointly learning class labels and semantic embedding [10].”
BackgroundMethodExtreme Reverse Projection Learning for Zero-Shot Recognition
Guan, J., Zhao, A., & Lu, Z. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-20887-5_8 · Cited by 5
Few-Shot Zero-Shot Learning: Knowledge Transfer with Less Supervision
Fei, N., Guan, J., Lu, Z., & Gao, Y. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-69535-4_36 · Cited by 5
Zero-Shot Action Recognition with Three-Stream Graph Convolutional Networks
Wu, N., & Kawamoto, K. (2021) · Sensors · doi:10.3390/s21113793 · Cited by 4
“Early ZSL methods [23–25] used attributes to represent the categories.”
MethodVocabulary-informed Extreme Value LearningPreprint
Fu, Y., Dong, H., Ma, Y.-F., Zhang, Z., & Xue, X. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1705.09887 · Cited by 4
“Overfeat features are used for AwA provided in [12].”
MethodSelf-citationNovelty detection using ensembles with regularized disagreement.Preprint
Ţifrea, A., Stavarache, E., & Yang, F. (2020) · arXiv (Cornell University) · Cited by 4
Class-wise and instance-wise contrastive learning for zero-shot learning based on VAEGAN
Zheng, B., Li, Z., & Li, J. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.126671 · Cited by 3
SEER-ZSL: Semantic Encoder-Enhanced Representations for Generalized Zero-Shot Learning
Heyden, W., Ullah, H., Siddiqui, M. S., & Machot, F. A. (2025) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw65960.2025.00009 · Cited by 3
“Alternative approaches to align seen and unseen data are presented in [22, 65].”
Ao, X., Li, X.-H., Zhang, X.-Y., & Liu, C. (2025) · Pattern Recognition · doi:10.1016/j.patcog.2025.112251 · Cited by 3
PNMT: Zero-Resource Machine Translation with Pivot-Based Feature Converter
Li, L., Hu, W., & Luo, M. (2025) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2025.064349 · Cited by 3
Boosting Zero-Shot Image Classification via Pairwise Relationship Learning
Li, H., Wu, H., Lin, S., Lin, L., Luo, X., & Ebroul Izquierdo, M. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-54181-5_6 · Cited by 3
Co-Learning for Few-Shot Learning
Xu, R., Xing, L., Shao, S., Liu, B., Zhang, K., & Liu, W. (2022) · Neural Processing Letters · doi:10.1007/s11063-022-10770-4 · Cited by 6
“2 Semi-Supervised Learning There is a lot of reference on semi-supervised learning [3], and the algorithms most relevant to our work are self-training [31,47], transductive learning [8,15].”
BackgroundTao, S., Tian, Z., Bai, L., Xu, Y., Kuang, C., & Liu, X. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.107711 · Cited by 3
Guo, J., & Guo, S. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00021 · Cited by 3
“from the external resource [9, 8], they may have some limitations to adapt to more substantial and new scenarios with increasing classes in the real-world situation.”
BackgroundMethodInfluentialCross-domain mapping learning for transductive zero-shot learning
Ding, M., Wang, Z., & Lu, Z. (2019) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2019.07.004 · Cited by 3
Zero-Shot Learning Using Stacked Autoencoder with Manifold Regularizations
Song, J., Shi, G., Xie, X., & Gao, D. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803509 · Cited by 3
“This phenomenon is often called the domain shift problem [9].”
BackgroundZero-Shot Visual Recognition via Semantic Attention-Based Compare Network
Nian, F., Sheng, Y., Wang, J., & Li, T. (2020) · IEEE Access · doi:10.1109/access.2020.2971174 · Cited by 3
Zero-Shot Embedding via Regularization-Based Recollection and Residual Familiarity Processes
Lyu, M., Han, H., & Bai, X. (2021) · IEEE Transactions on Systems Man and Cybernetics Systems · doi:10.1109/tsmc.2021.3102834 · Cited by 3
“D word2vec embeddings of AwA2 class names are provided by [35], which were obtained from a skip-gram model trained on a corpus of 4.6M Wikipedia articles.”
BackgroundMethodInfluentialDynamic Knowledge Distillation with Cross-Modality Knowledge Transfer
Wang, G. (2021) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3474085.3481034 · Cited by 3
ZS3D-Net: Zero-Shot Classification Network for 3D Models
Bai, J., Tao, Y., & Fan, Y. (2022) · Journal of Computer-Aided Design & Computer Graphics · doi:10.3724/sp.j.1089.2022.19173 · Cited by 3
Multi-label Generalized Zero-Shot Learning Using Identifiable Variational Autoencoders
Gull, M., & Arif, O. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-43404-4_3 · Cited by 3
Improving Class Representation for Zero-Shot Action Recognition
Zhou, L. M., & Mao, J. (2023) · ACM International Conference on Multimedia in Asia (MMAsia) · doi:10.1145/3595916.3626453 · Cited by 3
Encyclopedia enhanced semantic embedding for zero-shot learning
Jia, Z., Zhang, J., Huang, K., & Tan, T. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296489 · Cited by 3
“Many computer vision and machine learning methods, such as probabilistic models [8, 9, 10], canonical correlation analysis [11, 12], metric learning methods [13, 14] and graphical models [15] are exploited to solve the ZSL problem.”
MethodCoupled autoencoders learning for zero-shot classification with domain shift
Sun, G., Wu, S., Gao, G., Wu, F., & Jing, X. (2017) · IEEE International Conference on Progress in Informatics and Computing · doi:10.1109/pic.2017.8359516 · Cited by 3
Generating Manifold-Aligned Semantic Feature for Zero-Shot Learning
Wang, J., Li, Y., Pang, Z., & Wang, D. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451685 · Cited by 3
“To achieve this goal, one solution is to select some visual samples in X for each semantic embedding and average them to obtain the new one [10, 14, 15].”
BackgroundSampled in Pairs and Driven by Text: A New Graph Embedding Framework
Chen, L., Qu, Y., Wang, Z., Qiu, L., Zhang, W., Chen, K., Zhang, S., & Yu, Y. (2019) · International World Wide Web Conference (WWW) · doi:10.1145/3308558.3313520 · Cited by 4
“Although some previous works [13, 22, 52] discuss graph representation learning in zero-shot setting, they do not solve the same problem as we do.”
BackgroundA Zero-Shot based Fingerprint Presentation Attack Detection SystemPreprint
Liu, H., Zhang, W., Liu, G., & Liu, F. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.04908 · Cited by 3
“ZSL generally includes inductive ZSL [2,10] and transductive ZSL [16,41].”
BackgroundInfluentialSEMANTIC BORROWING FOR GENERALIZED ZERO-SHOT LEARNINGPreprint
Chen, X. (2021) · doi:10.36227/techrxiv.13683919.v1 · Cited by 3
Zero-sample surface defect detection and classification based on semantic feedback neural networkPreprint
Guo, Y., Fan, Y., Xiang, Z., Wang, H., Meng, W., & Xu, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.07959 · Cited by 5
“Pig tails are very different in appearance compared to tigers, zebras and other animals), which may lead to the problem of domain shift[31], so that when the model recognizes the unseen class, it is often classified into the seen class, resulting in a decrease in model accuracy.”
BackgroundMulti-modal scene categorization using multi-tasks learning
Peng, X., Li, Y., Jianhua, L., Xu, J., & Lu, Y. (2016) · International Conference on Signal Processing · doi:10.1109/icsp.2016.7878000 · Cited by 2
“[19] used CCA to fuse the multi-modal image feature, such as co-occurrence matrix, Zernike Moments.”
MethodA Simple and Explainable Method for Uncertainty Estimation using Attribute Prototype Networks
Zelenka, C., Göhring, A., Kazempour, D., Hünemörder, M., Schmarje, L., & Kröger, P. (2023) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw60793.2023.00491 · Cited by 4
“This approach has a wide range of applications, particularly in areas such as object and activity recognition, where it is unrealistic to obtain labeled data for all possible classes [29].”
BackgroundKong, Z., Gao, N., Zhang, Y., & Liu, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-53305-1_7 · Cited by 2
Bridging the Projection Gap: Overcoming Projection Bias Through Parameterized Distance Learning
Zhang, C., Jin, M., Yu, Q., Xue, H., Gowda, S. N., & Jin, X. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-0966-6_15 · Cited by 2
“Consequently, without appropriate adjustments to the embedding space for unseen classes, a problem of projection domain bias may arise Fu et al. [2015a], Zhao et al. [2017], Jia et al. [2019].”
BackgroundGenerate, Transduct, Adapt: Iterative Transduction with VLMs
Saha, O., Lawrence, L., Horn, G. V., & Maji, S. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00135 · Cited by 7
“Access to the entire unlabeled test set enables inference through methods such as label propagation [10, 50], clustering [48, 50, 59], among others [4, 19].”
Enhancing Zero-Shot Learning Through Kernelized Visual Prototypes and Similarity Learning
Cheng, K., & Fang, B. W. (2025) · Mathematics · doi:10.3390/math13030412 · Cited by 2
Learning Generalizable Contrastive Representations for Graph Zero-Shot Learning
Yi, S., Mao, Z., Zheng, K., Xiao, Z., Qiao, Z., Chen, C., Hua, X., Zhou, Y., Zhang, M., & Ju, W. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2025.3599043 · Cited by 4
“As such, the correlative relationship across different classes is established and the cross-class transferability can be well enhanced, which can alleviate the domain-shift issue [22].”
Zero-shot Learning with Many Classes by High-rank Deep Embedding Networks
Guo, Y., Ding, G., Han, J., Shao, H., Lou, X., & Dai, Q. (2019) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/337 · Cited by 2
Latent multi-view semi-supervised classification by using graph learning
Huang, Y., Yuan, H., & Lai, L. L. (2020) · International Journal of Wavelets Multiresolution and Information Processing · doi:10.1142/s0219691320500393 · Cited by 2
Exploring the Limits of Zero-Shot Learning - How Low Can You Go?
Dandu, H., Sharma, K., & Bhandarkar, S. M. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00187 · Cited by 2
“Although widely cited, DAP and IAP suffer from domain shift [8] where the intermediate functions learned from the auxiliary information are observed to introduce an unknown bias in the absence of adaptation to the target domain.”
BackgroundA Constrained Generative Approach to Generalized Zero-shot Object Recognition
Das, D., & Lee, C. S. G. (2021) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc52423.2021.9659235 · Cited by 5
“Popular transductive approaches mostly include using multiple sources [16], dictionary learning [17] with shared bases and semi-supervised learning [18].”
Background
Metadata
- Authors
- Fu, Y., Hospedales, T. M., Xiang, T., & Gong, S.
- Affiliations
- Walt Disney (United States) (US)
- Queen Mary University of London (GB)
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2015-11-01
- Volume
- 37
- Issue
- 11
- Pages
- 2332-2345
- ISSN
- 0162-8828 (ISSN-L)
- 2160-9292
- OpenAlex
- W2141350700
- Cited by
- 514
- References
- 74
- Citations by year
- 2026: 17 · 2025: 32 · 2024: 32 · 2023: 52 · 2022: 60 · 2021: 75 · 2020: 62 · 2019: 57 · 2018: 64 · 2017: 41 · 2016: 21 · 2015: 1
- FWCI
- 41.51
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Transactions on Pattern Analysis and Machine Intelligence (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Queen Mary Research Online (Queen Mary University of London) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Keywords
- zero-shot learning, transductive learning, multi-view embedding, semantic representation, transfer learning, manifold structure
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.