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 201–300 of 512
Improving CNN Classifiers by Estimating Test-Time Priors
Šulc, M., & Matas, J. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00402 · Cited by 20
“In transductive learning [6], the source and target tasks are the same, but the data representations or distributions are different.”
BackgroundTransductive Zero-Shot Hashing for Multilabel Image Retrieval
Zou, Q., Cao, L., Zhang, Z., Chen, L. F., & Wang, S. (2020) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.3043298 · Cited by 22
“The goal of ZSH [39], [40] is to transfer the model trained on the seen data to unseen data via other available knowledge, such as word vector representations.”
Learning unseen visual prototypes for zero-shot classification
Li, X., Fang, M., Feng, D., Li, H., & Wu, J. (2018) · Knowledge-Based Systems · doi:10.1016/j.knosys.2018.06.034 · Cited by 19
Fuzzy Multioutput Transfer Learning for Regression
Che, X., Zuo, H., Lü, J., & Chen, D. (2021) · IEEE Transactions on Fuzzy Systems · doi:10.1109/tfuzz.2021.3083956 · Cited by 19
“…[14] (e.g., using the previously classified French news to identify the classification of English news), computer image and vision processing [15] (e.g., constructing a recognition model to capture novel visual categories), and Wi-Fi location estimation [16] (e.g., discovering users’ current…”
BackgroundAn Iterative Co-Training Transductive Framework for Zero Shot Learning
Liu, B., Hu, L., Dong, Q., & Hu, Z. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3100552 · Cited by 22
Fabric recognition using zero-shot learning
Wang, F., Liu, H., Sun, F., & Pan, H. (2019) · Tsinghua Science & Technology · doi:10.26599/tst.2018.9010095 · Cited by 20
Bidirectional generative transductive zero-shot learning
Li, X., Zhang, D., Ye, M., Li, X., Dou, Q., & Lv, Q. (2020) · Neural Computing and Applications · doi:10.1007/s00521-020-05322-7 · Cited by 20
“They are: transductive multi-view zero-shot learning (TMV) [10], shared model space (SMS) [16], quasi-fully supervised learning (QFSL) [37].”
BackgroundObject Priors for Classifying and Localizing Unseen Actions
Mettes, P., Thong, W. E., & Snoek, C. G. M. (2021) · International Journal of Computer Vision · doi:10.1007/s11263-021-01454-y · Cited by 24
“The attribute-based approach has been extended by using knowledge about test videodistributions in transductive settings (Fu et al. 2015;Xu et al. 2017) and by incorporating domain adaptation (Kodirov et al.”
A zero-shot learning framework via cluster-prototype matching
Zhang, J., Li, Q., Geng, Y., Wang, W., Sun, W., Shi, C., & Ding, Z. (2021) · Pattern Recognition · doi:10.1016/j.patcog.2021.108469 · Cited by 19
A Generalized Zero-Shot Learning Framework for PolSAR Land Cover Classification
Gui, R., Xu, X., Wang, L., Yang, R. Q., & Pu, F. (2018) · Remote Sensing · doi:10.3390/rs10081307 · Cited by 20
“Since ZSL and GZSL depend on the semantic relevance to identify unseen classes, if the semantic relevance between the seen classes and unseen classes are poor, the above-mentioned domain shift problem [26,51] tends to appear, making it difficult to identify the unseen classes with poor semantic relevance accurately.”
BackgroundClass-Prototype Discriminative Network for Generalized Zero-Shot Learning
Huang, S., Lin, J., & Huangfu, L. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.2968213 · Cited by 19
“Such connections can be the mappings of visual → semantic [7], [13]–[15], semantic → visual [16], [17], and visual → intermediate representation ← semantic [18]–[21].”
BackgroundZhao, B., Sun, X., Fu, Y., Yao, Y., & Wang, Y. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.04360 · Cited by 19
“The label embedding of each class is a one-hot vector in FSL, while we use auxiliary knowledge Lampert et al. (2013); Fu et al. (2015b) to embed each label to be a semantic vector in ZSL.”
BackgroundMethodSelf-citationA Comprehensive Survey of Convolutions in Deep Learning: Applications, Challenges, and Future TrendsPreprint
Younesi, A., Ansari, M., Fazli, M., Ejlali, A., Shafique, M., & Henkel, J. (2024) · Repository KITopen (Karlsruhe Institute of Technology) · doi:10.48550/arxiv.2402.15490 · Cited by 21
“Zero-shot learning emerges as a promising area where CNNs imagine possibilities beyond the limitations of labeled data [370], [371], [372], [373], [374].”
BackgroundGAN-MVAE: a discriminative latent feature generation framework for generalized zero-shot learning
Ma, P., Lu, H., Yang, B., & Ran, W. (2022) · Pattern Recognition Letters · doi:10.1016/j.patrec.2022.02.002 · Cited by 18
Yang, G., Ye, Z., Zhang, R., & Huang, K. (2022) · Applied Computing and Intelligence · doi:10.3934/aci.2022001 · Cited by 20
Dual-focus transfer network for zero-shot learning
Jia, Z., Zhang, Z., Shan, C., Wang, L., & Tan, T. N. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.126264 · Cited by 20
Radar Active Jamming Recognition under Open World Setting
Zhang, Y., Zhao, Z., & Bu, Y. (2023) · Remote Sensing · doi:10.3390/rs15164107 · Cited by 18
Deep Prototypical Networks for Imbalanced Time Series Classification under Data Scarcity
Huang, C., Wu, X., Zhang, X., Lin, S., & Chawla, N. V. (2019) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3357384.3358162 · Cited by 18
“For example, in enterprise or healthcare scenarios, labelling samples is often very expensive or even impossible [7, 14], yet it is important to build effective time series classification models given only a handful of labeled examples per class.”
BackgroundLearning Joint Feature Adaptation for Zero-Shot RecognitionPreprint
Zhang, Z., & Saligrama, V. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.07593 · Cited by 19
“n to incorporate test-time unseen-class data in target domain into ZSL/ZSR as unlabeled data analogous to the transductive setting. This has led to approaches that attempt to account for domain shift [25,17,18,21,58]. In this setting, during test time, we are given a list of all unlabelled target instances in addition…”
BackgroundMethodInfluentialA Joint Generative Model for Zero-Shot Learning
Gao, R., Hou, X., Qin, J., Liu, L., Zhu, F., & Zhang, Z. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11018-5_50 · Cited by 18
Convolutional prototype learning for zero-shot recognition
Liu, Z., Zhang, X., Zhu, Z., Zheng, S., Zhao, Y., & Cheng, J. (2020) · Image and Vision Computing · doi:10.1016/j.imavis.2020.103924 · Cited by 19
Cross-class generative network for zero-shot learning
Liu, J., Zhang, Z., & Yang, G. (2020) · Information Sciences · doi:10.1016/j.ins.2020.12.063 · Cited by 17
A semi-supervised zero-shot image classification method based on soft-target
Ji, Z., Wang, Q., Cui, B., Pang, Y., Cao, X., & Li, X. (2021) · Neural Networks · doi:10.1016/j.neunet.2021.05.019 · Cited by 18
Learning Prototype via Placeholder for Zero-shot Recognition
Yang, Z., Liu, Y., Xu, W., Huang, C., Zhou, L., & Tong, C. (2022) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2022/217 · Cited by 17
“The problem of domain shift in ZSL is proposed by [Fu et al., 2015] and known as the projection domain shift.”
BackgroundInfluentialSimilarity preserving feature generating networks for zero-shot learning
Ma, Y., Xu, X., Shen, F., & Shen, H. T. (2020) · Neurocomputing · doi:10.1016/j.neucom.2019.08.111 · Cited by 19
Transductive Visual-Semantic Embedding for Zero-shot Learning
Xu, X., Shen, F., Yang, Y., Shao, J., & Huang, Z. H. (2017) · International Conference on Multimedia Retrieval · doi:10.1145/3078971.3078977 · Cited by 16
“The second problem is the inherent projection shift problem, which has also been addressed in [9, 30].”
BackgroundMethodInfluentialA decadal survey of zero-shot image classification
Pang, Y., Wang, H., Yu, Y., & Ji, Z. (2019) · Scientia Sinica Informationis · doi:10.1360/n112018-00312 · Cited by 21
“主要用两种思路缓解域偏移问题 : 一类是基于直推式学习的方法 [47,94,95] , 主要思路是通过挖掘未见类 样本的流形信息缓解域偏移的程度 .”
Deep Matching AutoencodersPreprint
Mukherjee, T., Yamada, M., & Hospedales, T. M. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.06047 · Cited by 19
“However it has a few crucial benefits: (i) It can be learned in a semi-supervised way, which encompasses the transductive [14, 43] and semi-supervised [43] variants of ZSL.”
BackgroundSelf-citationFrom Classical to Generalized Zero-Shot Learning: a Simple Adaptation ProcessPreprint
Cacheux, Y. L., Borgne, H. L., & Crucianu, M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.10120 · Cited by 17
Federated zero-shot learning with mid-level semantic knowledge transfer
Sun, S., Si, C., Wu, G., & Gong, S. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110824 · Cited by 18
Self-citationVisual–Semantic Graph Matching Net for Zero-Shot Learning
Duan, B., Chen, S., Guo, Y., Xie, G.-S., Ding, W., & Wang, Y. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2024.3499377 · Cited by 16
Zero shot learning based on class visual prototypes and semantic consistency
Li, X., Fang, M., Li, H., & Wu, J. (2020) · Pattern Recognition Letters · doi:10.1016/j.patrec.2020.04.029 · Cited by 15
Explainable Neural Network-based Modulation Classification via Concept Bottleneck Models
Wong, L. J., & McPherson, S. (2021) · IEEE Annual Computing and Communication Workshop and Conference (CCWC) · doi:10.1109/ccwc51732.2021.9376108 · Cited by 19
“…as well, where the histogram of the Independent Bottleneck model order regression-head contains all five peaks, while the histogram of the Joint Bottleneck model order regression-head only contains four peaks and has created a single output range for the higher order modulation schemes: 16 and 64.”
BackgroundResultWu, H., Wang, Z., Qing, F., & Li, S. (2021) · Electronics · doi:10.3390/electronics10030270 · Cited by 15
“[5]) when translate domain-specific text.”
Learning from Graph Propagation via Ordinal Distillation for One-Shot Automated Essay Scoring
Jiang, Z., Liu, M., Yin, Y., Yu, H., Cheng, Z., & Gu, Q. (2021) · International World Wide Web Conference (WWW) · doi:10.1145/3442381.3450017 · Cited by 15
“The classical methods include the Transductive Support Vector Machine (TSVM) [15] and the graph-based label propagation [12, 52, 53].”
MethodLiu, Z., Yang, R., Liu, W., & Liu, X. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.126877 · Cited by 15
A Transferable Generative Framework for Multi-Label Zero-Shot Learning
Ma, P., He, Z., Ran, W., & Lu, H. (2023) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2023.3324648 · Cited by 15
“However, since the seen classes and the unseen classes are disjoint, the embedding model learned only on the seen classes will produce a large bias when directly used for unseen classes prediction ( i.e. , projection domain shifts [38]).”
BackgroundDeep Semisupervised Zero-Shot Learning with Maximum Mean Discrepancy
Zhang, L., Liu, J., Luo, M., Chang, X., & Zheng, Q. (2018) · Neural Computation · doi:10.1162/neco_a_01071 · Cited by 15
Zero-shot learning with regularized cross-modality ranking
Yu, Y., Ji, Z., Guo, J., & Pang, Y. (2017) · Neurocomputing · doi:10.1016/j.neucom.2016.06.085 · Cited by 14
“Side information CNN Feature AwA CUB AMP [21] A+W Overfeat 66.00 - TMV-HLP [22] A+W Overfeat 73.50 47.90 SJE [23] A+W+H GoogleNet 73.50 51.70 ReCMR A+W VGG 79.72 56.03 Table 5: Running time comparison.”
BackgroundMethodInfluentialZero-shot learning with Multi-Battery Factor Analysis
Ji, Z., Xie, Y., Pang, Y., Chen, L., & Zhang, Z. M. (2017) · Signal Processing · doi:10.1016/j.sigpro.2017.03.023 · Cited by 15
“However, although multiple types of side information are utilized, they still exploit each type of side information in its own semantic space independently, and then just combine the predicted scores together [13], [16].”
BackgroundMethodLearning domain invariant unseen features for generalized zero-shot classification
Li, X., Fang, M., Li, H., & Wu, J. (2020) · Knowledge-Based Systems · doi:10.1016/j.knosys.2020.106378 · Cited by 15
Bidirectional Mapping Coupled GAN for Generalized Zero-Shot Learning
Shermin, T., Teng, S. W., Sohel, F. A., Murshed, M., & Lu, G. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3135480 · Cited by 16
Adaptive Few-Shot Learning Algorithm for Rare Sound Event Detection
Zhao, C., Wang, J., Li, L., Qu, X., & Xiao, J. Z. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892604 · Cited by 14
A Latent Representation Generalizing Network for Domain Generalization in Cross-Scenario Monitoring
He, S., Chen, F., & Chen, H. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3296942 · Cited by 18
“The distributions of training and testing data are required to be consistent [20], [21].”
BackgroundTask-like training paradigm in CLIP for zero-shot sketch-based image retrieval
Zhang, H., Cheng, D., Jiang, H., Liu, J., & Kou, Q. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-17675-x · Cited by 16
“Accordingly, the inherent modality gap between sketches and images [10, 11], the semantic gap between unseen and seen classes [12–14], are inescapable.”
Zero-Shot Learning via Low-Rank-Representation Based Manifold Regularization
Meng, M., & Zhan, X. (2018) · IEEE Signal Processing Letters · doi:10.1109/lsp.2018.2857201 · Cited by 14
Database of Gesture Attributes: Zero Shot Learning for Gesture Recognition
Madapana, N., & Wachs, J. P. (2019) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg.2019.8756548 · Cited by 14
Semantic softmax loss for zero-shot learning
Ji, Z., Sun, Y., Yu, Y., Guo, J., & Pang, Y. (2018) · Neurocomputing · doi:10.1016/j.neucom.2018.08.014 · Cited by 14
Zhou, Q., Zhang, R., Zhang, F., & Jing, X. (2022) · EURASIP Journal on Wireless Communications and Networking · doi:10.1186/s13638-022-02099-2 · Cited by 14
“For the classes with entirely missing data, the zero-sample learning scheme is always considered by expert-set linguistic descriptions, transfer learning, or matrix transformations [32, 33].”
MethodTransductive zero-shot learning with generative model-driven structure alignment
Liu, Y., Tao, K., Tian, T., Gao, X., Han, J., & Shao, L. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110561 · Cited by 13
Zero-Shot Detection of Visual Food Safety Hazards via Knowledge-Enhanced Feature Synthesis
Guo, L., Hu, X., Liu, W., & Liu, Y. (2025) · Applied Sciences · doi:10.3390/app15116338 · Cited by 36
Fs-DSM: Few-Shot Diagram-Sentence Matching via Cross-Modal Attention Graph Model
Hu, X., Zhang, L., Liu, J., Zheng, Q., & Zhou, J. J. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3112294 · Cited by 13
Ma, S., Liu, C., Li, Z., & Yang, W. L. (2022) · Remote Sensing · doi:10.3390/rs14184533 · Cited by 13
Webly Supervised Semantic Embeddings for Large Scale Zero-Shot Learning
Cacheux, Y. L., Popescu, A., & Borgne, H. L. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-69544-6_31 · Cited by 13
Domain-aware multi-modality fusion network for generalized zero-shot learning
Wang, J., Wang, X., & Zhang, H. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.02.056 · Cited by 13
Learning to Infer Unseen Single-/ Multi-Attribute-Object Compositions With Graph Networks
Chen, H., Jiang, J., & Zheng, N. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3273712 · Cited by 15
“As a result, these methods do not achieve satisfactory performance and suffer from the “domain shift” problem [45] - the distribution of the testing data is different from that of the training data.”
From Classical to Generalized Zero-Shot Learning: A Simple Adaptation Process
Cacheux, Y. L., Borgne, H. L., & Crucianu, M. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-05716-9_38 · Cited by 18
Fusion by synthesizing: A multi-view deep neural network for zero-shot recognition
Xu, X., Zhou, X. S., Shen, F., Gao, L., Shen, H. T., & Li, X. (2019) · Signal Processing · doi:10.1016/j.sigpro.2019.05.022 · Cited by 13
Superclass-aware visual feature disentangling for generalized zero-shot learning
Niu, C., Shang, J., Zhou, Z., & Yang, J. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.125150 · Cited by 13
Zero-Shot Predicate Prediction for Scene Graph Parsing
Li, Y., Yang, X., Huang, X., Ma, Z., & Xu, C. (2022) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2022.3155928 · Cited by 16
ATZSL: Defensive Zero-Shot Recognition in the Presence of Adversaries
Zhang, X., Gui, S., Jin, J., Zhu, Z., & Zhao, Y. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3258624 · Cited by 15
“According to whether information about the testing data is involved during model learning, existing ZSL models consist of inductive [9], [23], [24], [25], [26] and transductive settings [27], [28], [29], [30], [31].”
BackgroundHao, Y., Su, Y., Lin, G., Su, H., & Wu, Q. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.109843 · Cited by 16
Dual-View Ranking with Hardness Assessment for Zero-Shot Learning
Guo, Y., Ding, G., Han, J., Ding, X., Zhao, S., Wang, Z., Yan, C. C., & Dai, Q. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33018360 · Cited by 13
“Since our work focuses on inductive ZSL, some transductive ZSL approaches (Fu et al. 2015a; Kodirov et al. 2015; Guo et al. 2016) are not used as baseline approaches.”
BackgroundMethodInductive Zero-Shot Image Annotation via Embedding Graph
Wang, F., Liu, J., Zhang, S., Zhang, G., Li, Y., & Yuan, F. (2019) · IEEE Access · doi:10.1109/access.2019.2925383 · Cited by 12
Class label autoencoder with structure refinement for zero-shot learning
Lin, G., Fan, C., Chen, W., Chen, Y., & Zhao, F. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.11.061 · Cited by 12
Generative Adversarial Zero-shot Learning via Knowledge GraphsPreprint
Geng, Y., Chen, J., Chen, Z., Ye, Z., Yuan, Z., Jia, Y., & Chen, H. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.03109 · Cited by 12
“…to generate samples for class zebra , another significantly different class tiger which is also annotated with “stripe” may also obtain synthesized samples with similar features as zebra (i.e., domain-shift problem [6]), especially when tiger lacks of other representative attribute annotations.”
BackgroundLearning Multipart Attention Neural Network for Zero-Shot Classification
Meng, M., Wei, J., & Wu, J. (2020) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2020.3044313 · Cited by 11
“Most of the state-of-the-art ZSL methods [13]–[21] use deep CNN features for visual feature representation which are extracted with pretrained CNN models.”
BackgroundMethodZhang, C., Yuan, M., Ma, X., Wei, P., Su, Y., Li, L., & Liu, Y. (2024) · IEEE Open Journal of Intelligent Transportation Systems · doi:10.1109/ojits.2024.3418552 · Cited by 11
“This learned shared space can be interpretable [44] or just an implicit representation space [45], [46].”
BackgroundMethodResultZero-Shot Recognition via Optimal Transport
Wang, W., Xu, H., Wang, G., Wang, W., & Carin, L. (2021) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv48630.2021.00351 · Cited by 12
“All these types of methods have been widely used in the standard and trans-ductive ZSL problems [19, 35, 46].”
MethodHandling new target classes in semantic segmentation with domain adaptation
Bucher, M., Vu, T.-H., Cord, M., & Pérez, P. (2021) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2021.103258 · Cited by 15
Fine-Grained Feature Generation for Generalized Zero-Shot Video Classification
Hong, M., Zhang, X., Li, G., & Huang, Q. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3247167 · Cited by 14
“Zero-shot learning is firstly studied by exploring visual attributes as semantic descriptions [14], [15], [16], [17], [18], [19], [20].”
MethodZero-shot pipeline fault detection using percussion method and multi-attribute learning model
Peng, L., Huang, W., Zhang, J., & Du, G. (2025) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2025.112427 · Cited by 10
Visual-guided attentive attributes embedding for zero-shot learning
Zhang, R., Zhu, Q., Xu, X., Zhang, D., & Huang, S.-J. (2021) · Neural Networks · doi:10.1016/j.neunet.2021.07.031 · Cited by 10
Hard Zero Shot Learning for Gesture Recognition
Madapana, N., & Wachs, J. P. (2018) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2018.8545869 · Cited by 12
Wheel Hub Defects Image Recognition Base on Zero-Shot Learning
Sun, X., Gu, J., Wang, M., Meng, Y., & Shi, H. (2021) · Applied Sciences · doi:10.3390/app11041529 · Cited by 10
“Fu et al. [38] used multiple semantic information sources to explore the projection domain shift and implemented label transfer for unlabeled data of unseen classes to solve this problem.”
Towards Recognizing Unseen Categories in Unseen Domains
Mancini, M., Akata, Z., Ricci, E., & Caputo, B. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58592-1_28 · Cited by 9
MeshCLIP: Efficient cross-modal information processing for 3D mesh data in zero/few-shot learning
Song, Y., Liang, N., Guo, Q., Dai, J., Bai, J., & He, F. (2023) · Information Processing & Management · doi:10.1016/j.ipm.2023.103497 · Cited by 11
Transductive Zero-Shot Hashing via Coarse-to-Fine Similarity Mining
Lai, H. (2018) · International Conference on Multimedia Retrieval · doi:10.1145/3206025.3206026 · Cited by 9
“utions of the seen classes and the novel classes are different, the projection functions learned by the seen classes without any adaptation to the novel classes may cause data bias problem. Fu et al. [1] proposed transductive multi-view embedding to solve the projection domain shift problem. Guo et al. [28] proposed t…”
BackgroundMethodEnhanced VAEGAN: a zero-shot image classification method
Ding, B., Fan, Y., He, Y., & Zhao, J. (2022) · Applied Intelligence · doi:10.1007/s10489-022-03869-7 · Cited by 9
“The VAEGAN possesses advantages of the VAE and GAN, which can effectively alleviate the problems of hubness [12, 13] and domain shift [14, 15].”
BackgroundMulti-View Deep Gaussian Processes for Supervised Learning
Dong, W., & Sun, S. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3316671 · Cited by 10
“Existing research has shown that multi-view learning can obtain better learning performance than single view learning through mutualcollaborationandpromotionbetweenviews[8],[9],[10], [11],[12].”
BackgroundM3R: Masked Token Mixup and Cross-Modal Reconstruction for Zero-Shot Learning
Zhao, P., Wang, Q., & Yin, Y. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3612104 · Cited by 11
Generalizability and robustness evaluation of attribute-based zero-shot learning
Rossi, L., Fiorentino, M. C., Mancini, A., Paolanti, M., Rosati, R., & Zingaretti, P. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106278 · Cited by 9
Zero-Shot Crowd Behavior Recognition
Xu, X., Gong, S., & Hospedales, T. M. (2017) · Elsevier eBooks · doi:10.1016/b978-0-12-809276-7.00018-7 · Cited by 9
Self-citationAdaptive Adjustment with Semantic Feature Space for Zero-shot Recognition
Guo, J., & Guo, S. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8682869 · Cited by 9
Multi-Label Zero-Shot Learning With Transfer-Aware Label Embedding Projection
Ye, M., & Guo, Y. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803720 · Cited by 9
“One primary source deployed in ZSL for bridging the gap between seen and unseen classes is the attribute description of the class labels [5, 10, 7, 11].”
BackgroundCross-Layer Autoencoder for Zero-Shot Learning
Zhang, Z., Li, Y., Yang, J. K., Li, Y., & Gao, M. (2019) · IEEE Access · doi:10.1109/access.2019.2953454 · Cited by 10
A comprehensive review on zero-shot-learning techniques
Lazaros, K., Koumadorakis, D. E., Vrahatis, A. G., & Kotsiantis, S. B. (2024) · Intelligent Decision Technologies · doi:10.3233/idt-24027 · Cited by 8
Dynamic VAEs via semantic-aligned matching for continual zero-shot learning
Yang, J., Hu, B., Li, H., Liu, Y., Gao, X., Han, J., Chen, F., & Wu, X. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.111199 · Cited by 10
Hirt, R., Kühl, N., Peker, Y., & Satzger, G. (2020) · IEEE Conference on Business Informatics · doi:10.1109/cbi49978.2020.00010 · Cited by 9
“This type of prediction problem is called “zero shot” [12] and is especially relevant for branches without historic data available, e.g., new branches or branches in planning.”
Feature Selection for Zero-Shot Gesture Recognition
Madapana, N., & Wachs, J. P. (2020) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg47880.2020.00046 · Cited by 8
“Hence, it has been a hot topic in the machine learning community with a plethora of academic articles published every year pushing the limits of unseen class accuracies [12]–[14].”
BackgroundKernelized distance learning for zero-shot recognition
Zarei, M. R., Taheri, M., & Long, Y. (2021) · Information Sciences · doi:10.1016/j.ins.2021.09.032 · Cited by 8
Region Semantically Aligned Network for Zero-Shot Learning
Wang, Z., Gou, Y., Li, J., Zhang, Y., & Yang, Y. (2021) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3459637.3482471 · Cited by 10
“Moreover, the problem of domain shift [15, 21, 64] further degrades the performance of the image encoder.”
Recognizing unknown objects with attributes relationship model
Hoo, W. L., & Chan, C. S. (2015) · Expert Systems with Applications · doi:10.1016/j.eswa.2015.07.049 · Cited by 8
Domain-aware Stacked AutoEncoders for zero-shot learning
Song, J., Shi, G., Xie, X., Wu, Q., & Zhang, M. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.12.017 · Cited by 8
Attribute-Induced Bias Eliminating for Transductive Zero-Shot Learning
Yao, H., Min, S., Zhang, Y., & Xu, C. (2021) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2021.3074252 · Cited by 10
Learning Class-Transductive Intent Representations for Zero-shot Intent Detection
Si, Q., Liu, Y., Fu, P., Lin, Z., Li, J., & Wang, W. (2021) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2021/540 · Cited by 11
“In GZSID, existing ZSL models usually suffer from the domain shift [Fu et al., 2015a] problem, in which utterances from unseen intents are almost always mistakenly classified into seen intents.”
MethodLearning the Compositional Domains for Generalized Zero-shot Learning
Dong, H., Fu, Y., Hwang, S. J., Sigal, L., & Xue, X. (2022) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2022.103454 · Cited by 10
Self-citationMoreu, E., Arazo, E., McGuinness, K., & O'Connor, N. E. (2022) · Expert Systems · doi:10.1111/exsy.13137 · Cited by 12
“This is a common setup in zero-shot learning (Fu et al., 2015; Kodirov et al., 2015; Song et al., 2018; Wan et al., 2019) that explores the performance of a model when the unlabelled target data, the test set in our case, is available during training.”
BackgroundZero-Shot Learning for Gesture Recognition
Madapana, N. (2020) · INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION · doi:10.1145/3382507.3421161 · Cited by 9
“Hence, it has been a hot topic in the machine learning community with a plethora of academic articles published every year pushing the limits of unseen class accuracies [4, 15, 20].”
BackgroundDisentangled features with direct sum decomposition for zero shot learning
Li, B., Han, C., Guo, T., & Zhao, T. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.09.065 · Cited by 10
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.