Recent Advances in Transfer Learning for Cross-Dataset Visual Recognition: A Problem-Oriented Perspective
The DOI matches its Crossref record; 41 of the 41 references it cites were found in a registry or an index.
Abstract
This article takes a problem-oriented perspective and presents a comprehensive review of transfer-learning methods, both shallow and deep, for cross-dataset visual recognition. Specifically, it categorises the cross-dataset recognition into 17 problems based on a set of carefully chosen data and label attributes. Such a problem-oriented taxonomy has allowed us to examine how different transfer-learning approaches tackle each problem and how well each problem has been researched to date. The comprehensive problem-oriented review of the advances in transfer learning with respect to the problem has not only revealed the challenges in transfer learning for visual recognition but also the problems (e.g., 8 of the 17 problems) that have been scarcely studied. This survey not only presents an up-to-date technical review for researchers but also a systematic approach and a reference for a machine-learning practitioner to categorise a real problem and to look up for a possible solution accordingly.
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
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333
Perceptual Losses for Real-Time Style Transfer and Super-Resolution
Johnson, J., Alahi, A., & Fei-Fei, L. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46475-6_43 · Cited by 11,283
How transferable are features in deep neural networks?
Yosinski, J., Clune, J., Bengio, Y., & Lipson, H. (2014) · arXiv · doi:10.48550/arxiv.1411.1792 · Cited by 6,061 · cited 5×
“Typical methods are Sparse coding [185], Low-rank representation [196], Deep Neural Networks [50, 189, 269], Stacked Denoising Auto-encoders (SDAs) [27, 77], and Attribute space [2, 124].”
BackgroundMethodInfluentialImage Style Transfer Using Convolutional Neural Networks
Gatys, L. A., Ecker, A. S., & Bethge, M. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.265 · Cited by 6,025
Adversarial Discriminative Domain Adaptation
Tzeng, E., Hoffman, J., Saenko, K., & Darrell, T. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.316 · Cited by 5,154 · cited 4×
“Motivated by adversarial learni g [84], the GAN-based domain adaptatio methods are proposed with the key idea that the JS divergence bet een dom ns are red ced [17, 70, 71, 226].”
BackgroundMethodInfluentialLearning Transferable Features with Deep Adaptation Networks
Long, M., Cao, Y., Wang, J., & Jordan, M. I. (2015) · arXiv · doi:10.48550/arxiv.1502.02791 · Cited by 4,505 · cited 5×
“Several example combinations are Correspondence and Higher-level representation [96], Higher-level representation and Statistic [147, 148, 243], Statistic and Geometric [273], Statistic and Self labelling [42], Correspondence and Class-based [46], Statistic and Class-based [52], and Higher-level representation and Cla…”
BackgroundMethodInfluentialDeCAF: 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 6×
“cture. Bengio [14] argue that more transferable features can be learned by deep networks since they are able to extract the unknown factors of variation that are intrinsic to the data. Donahue et al. [50] propose the deep convolutional representations named DeCAF, where a deep CNN model is pre-trained using the source…”
BackgroundMethodInfluentialDeep CORAL: Correlation Alignment for Deep Domain Adaptation
Sun, B., & Saenko, K. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-49409-8_35 · Cited by 3,757 · cited 4×
“For instance, Saenko et al [33] proposed a metric learning approach that could use labeled data for few categories from the target domain to predict the domain change for unlabeled target categories.”
MethodInfluentialA survey on concept drift adaptation
Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014) · ACM Computing Surveys · doi:10.1145/2523813 · Cited by 3,735
The class imbalance problem: A systematic study1
Japkowicz, N., & Stephen, S. (2002) · Intelligent Data Analysis · doi:10.3233/ida-2002-6504 · Cited by 3,295
NUS-WIDE: a real-world web image database from National University of Singapore
Chua, T.-S., Tang, J., Hong, R., Li, H., Luo, Z., & Zheng, Y. (2009) · Proceedings of the ACM International Conference on Image and Video Retrieval · doi:10.1145/1646396.1646452 · Cited by 3,025 · cited 3×
“ature of classic ation error across new domains [6, 4]. Along similar lines, there have been efforts focusing on domain shift issues for 2D object recognition applications. For instance, Saenko et al [33] proposed a metric learning approach that could use labeled data for few categories from the target domain to predi…”
BackgroundInfluentialEigen, D., & Fergus, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.304 · Cited by 2,862 · cited 3×
“Our solution is to pose this as a semi-supervised learning [57] problem: prototypes in each view are treated as labelled ‘instances’, and we exploit the manifold structure of the unlabelled data”
BackgroundMethodAdapting visual category models to new domains
Saenko, K., Kulis, B., Fritz, M., & Darrell, T. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-15561-1_16 · Cited by 2,847 · cited 4×
“ce and target datasets. Hence, the labelled data from each dataset are assumed to be available, whether sufficient or not. The commonly used methods include Feature augmentation [44], Metric learning [193], Linear Discriminative Model [264], and Bayesian Model [60]. Self Labelling: uses the source domain samples to tr…”
MethodInfluentialTorralba, A., & Efros, A. A. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995347 · Cited by 2,440 · cited 1×
“ned model to test data acquired from the scenarios different from those of the training data mainly due to domain divergence and task divergence [170]. This is particularly true in visual recognition [223] where external factors such as environments, lighting, background, sensor types, view angles, and post-processing…”
Deep domain confusion Maximizing for domain invariance
Tzeng, E., Hoffman, J., Zhang, N., Saenko, K., & Darrell, T. (2014) · arXiv · doi:10.48550/arxiv.1412.3474 · Cited by 2,285
Deep Hashing Network for Unsupervised Domain Adaptation
Venkateswara, H., Eusebio, J., Chakraborty, S., & Panchanathan, S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.572 · Cited by 2,171
Transfer Feature Learning with Joint Distribution Adaptation
Long, M., Wang, J., Ding, G., Sun, J., & Yu, P. S. (2013) · 2013 IEEE International Conference on Computer Vision · doi:10.1109/iccv.2013.274 · Cited by 2,071
Return of Frustratingly Easy Domain Adaptation
Sun, B., Feng, J., & Saenko, K. (2016) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v30i1.10306 · Cited by 2,004 · cited 1×
“tures that cannot be modelled by the source task. Recently, deep neural networks are applied for unsupervised domain adaptation. Sun and Saenko [2016] extend the CORrelation ALignment (CORAL) method [Sun et al. 2016] to learn a nonlinear transformation using deep neural network. Ganin and Lempitsky [2015] propose to m…”
MethodDescribing objects by their attributes
Farhadi, A., Endres, I., Hoiem, D., & Forsyth, D. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206772 · Cited by 1,917
Learning to detect unseen object classes by between-class attribute transfer
Lampert, C. H., Nickisch, H., & Harmeling, S. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206594 · Cited by 1,849 · cited 3×
“10 v.s. STL-10[63] # # ✓ RGB-D→Caltech256[25] ✓ # # # # ✓ Syn Signs v.s. GTSRB[70] # # ✓ NUS-WIDE[33] ✓ # # Wikipedia dataset[177] ✓ # # Pascal Sentence[186] ✓ # # MSCOCO[142] ✓ # # aP&Y[59] ✓ AwA[124] ✓ Caltech-UCSD CUB[231] ✓ Caltech-256[221] ✓ Car over time[92] ✓ STL-10 dataset[34] ✓ LabelMe →NUS-WIDE[235] ✓ Ou…”
MethodLearning to Discover Cross-Domain Relations with Generative Adversarial Networks
Kim, T., Cha, M., Kim, H., Lee, J. K., & Kim, J. (2017) · arXiv · doi:10.48550/arxiv.1703.05192 · Cited by 1,716
Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
Bousmalis, K., Silberman, N., Dohan, D., Erhan, D., & Krishnan, D. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.18 · Cited by 1,603 · cited 1×
“vergence is used to encourage samples from different domains to be non-discriminative with respect to domain labels [Tzeng et al. 2015; Ganin and Lempitsky 2015; Ganin et al. 2016; Tzeng et al. 2017; Bousmalis et al. 2017]. For example, Tzeng et al. [2015] proposed adding a domain classifier that predicts the binary do…”
MethodDataset Shift in Machine Learning
(2008) · doi:10.7551/mitpress/9780262170055.001.0001 · Cited by 1,553 · cited 1×
“d, where the previous and current tasks are the same. The extensive study has led to about a dozen of tutorial and survey papers published since 2009, from the analysis of the nature of dataset shift [184] to the formal definition and task-oriented categorization of transfer learning [170], and to the recent tutorial…”
BackgroundContinual Learning with Deep Generative Replay
Shin, H., Lee, J. K., Kim, J., & Kim, J. (2017) · arXiv · doi:10.48550/arxiv.1705.08690 · Cited by 1,385
The influence of improvement in one mental function upon the efficiency of other functions. (I).
Woodworth, R. S., & Thorndike, E. L. (1901) · Psychological Review · doi:10.1037/h0074898 · Cited by 1,223
Feature Generating Networks for Zero-Shot Learning
Xian, Y., Lorenz, T., Schiele, B., & Akata, Z. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00581 · Cited by 1,080
Zero-Shot Learning Through Cross-Modal Transfer
Socher, R., Ganjoo, M., Sridhar, H., Bastani, O., Manning, C. D., & Ng, A. Y. (2013) · arXiv · doi:10.48550/arxiv.1301.3666 · Cited by 997
Transfer Learning for Visual Categorization: A Survey
Shao, L., Zhu, F., & Li, X. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2014.2330900 · Cited by 895 · cited 1×
“he formal definition and task-oriented categorization of transfer learning [170], and to the recent tutorial and survey on deep learning based domain adaptation [37, 228]. Most of these survey papers [12, 35, 154, 157, 160, 173, 195, 214, 228, 245] are method-driven and provide up to the time a review of the evolution…”
MethodZero-Shot Learning — The Good, the Bad and the Ugly
Xian, Y., Schiele, B., & Akata, Z. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.328 · Cited by 891
Domain Generalization via Invariant Feature Representation
Muandet, K., Balduzzi, D., & Schölkopf, B. (2013) · arXiv · doi:10.48550/arxiv.1301.2115 · Cited by 771
Holistically-nested edge detection
Xie, S., & Tu, Z. (2017) · International Journal of Computer Vision · doi:10.1007/s11263-017-1004-z · Cited by 682
Domain Generalization for Object Recognition with Multi-task Autoencoders
Ghifary, M., Kleijn, W. B., Zhang, M., & Balduzzi, D. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.293 · Cited by 664
Li, Z., & Hoiem, D. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_37 · Cited by 659
Zero-Shot Learning via Semantic Similarity Embedding
Zhang, Z., & Saligrama, V. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.474 · Cited by 629 · cited 1×
“Hence, the second strategy is recognition by assuming a fixed transformation (W) between the attributes and the class labels [1, 3, 140, 181, 191, 248, 280, 281] to learn all attributes simultaneously: F (x ,y;W ) = θ (x)TWφ(y), where θ (x) and φ(y) represent image and class embeddings, both are given.”
BackgroundCross Modal Distillation for Supervision Transfer
Gupta, S., Hoffman, J., & Malik, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.309 · Cited by 530 · cited 4×
“tional neural networks are regularized using statistical regularization so that they have a shared representation that is invariant to different modalities. The paired correspondence data are used in [89], where a cross-modal supervision transfer method is proposed. The deep CNNs are pre-trained on the source data (e.…”
BackgroundInfluentialTransductive Multi-View Zero-Shot Learning
Fu, Y., Hospedales, T. M., Xiang, T., & Gong, S. (2015) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2015.2408354 · Cited by 514 · cited 4×
“(Figure used courtesy of [66])”
BackgroundInfluentialTransfer learning for activity recognition: a survey
Cook, D., Feuz, K. D., & Krishnan, N. C. (2013) · Knowledge and Information Systems · doi:10.1007/s10115-013-0665-3 · Cited by 447 · cited 3×
“type) solutions for this problem [35, 23, 11], Blitzer et al”
BackgroundMethodLearning with Augmented Features for Heterogeneous Domain Adaptation
Duan, L., Xu, D., & Tsang, I. (2012) · arXiv · doi:10.48550/arxiv.1206.4660 · Cited by 320
Self-citationLearning Aligned Cross-Modal Representations from Weakly Aligned Data
Castrejon, L., Aytar, Y., Vondrick, C., Pirsiavash, H., & Torralba, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.321 · Cited by 188 · cited 1×
“[23] introduce a cross-modal representation method across RGB modality, sketch modality, clipart, and textual descriptions of indoor scenes.”
MethodOnline domain adaptation of a pre-trained cascade of classifiers
Jain, V., & Learned-Miller, E. (2011) · CVPR 2011 · doi:10.1109/cvpr.2011.5995317 · Cited by 161 · cited 1×
“llow for discontinuities, but we leave this as future work. Fig. 6. Illustration of the continuous domain adaptation method [92] (Figure used courtesy of [92]) Self Labelling. Jain and Learned-Miller [99] address the online adaptation in the face detection task by adapting pre-trained classifiers using a Gaussian proc…”
Method
Cited by
Domain adaptation for medical image analysis: A survey
Guan, H., & Liu, M. (2022) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2021.3117407 · Cited by 1,057
“In a typical transfer learning setting, there are two concepts: “domain” and “task” [16]–[18].”
Synthetic Data for Deep Learning
Nikolenko, S. I. (2021) · Springer Optimization and Its Applications · doi:10.1007/978-3-030-75178-4 · Cited by 423
Deep Gait Recognition: A Survey
Sepas-Moghaddam, A., & Etemad, A. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3151865 · Cited by 307
GLOBEM: Cross-Dataset Generalization of Longitudinal Human Behavior Modeling
Xu, X., Liu, X., Zhang, H., Wang, W., Nepal, S., Sefidgar, Y., Seo, W., Kuehn, K. S., Huckins, J. F., Morris, M. E., et al. (2022) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3569485 · Cited by 96
Progressive modality cooperation for multi-modality domain adaptation
Zhang, W., Xu, D., Zhang, J., & Ouyang, W. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3052083 · Cited by 25
Self-citation
Metadata
- Authors
- Zhang, J., Li, W., Ogunbona, P., & Xu, D.
- Venue
- ACM Computing Surveys
- Published
- 2019-02-17
- Volume
- 52
- Issue
- 1
- Pages
- 1-38
- ISSN
- 0360-0300 (ISSN-L)
- 1557-7341
- OpenAlex
- W2916332638
- Cited by
- 85
- References
- 41
- FWCI
- 8.29
- Type
- journal article
- Language
- en
- Open access
- closed
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.