Transferable Curriculum for Weakly-Supervised Domain Adaptation
Open accessThe DOI matches its Crossref record; 13 of the 13 references it cites were found in a registry or an index.
Abstract
Domain adaptation improves a target task by knowledge transfer from a source domain with rich annotations. It is not uncommon that “source-domain engineering” becomes a cumbersome process in domain adaptation: the high-quality source domains highly related to the target domain are hardly available. Thus, weakly-supervised domain adaptation has been introduced to address this difficulty, where we can tolerate the source domain with noises in labels, features, or both. As such, for a particular target task, we simply collect the source domain with coarse labeling or corrupted data. In this paper, we try to address two entangled challenges of weaklysupervised domain adaptation: sample noises of the source domain and distribution shift across domains. To disentangle these challenges, a Transferable Curriculum Learning (TCL) approach is proposed to train the deep networks, guided by a transferable curriculum informing which of the source examples are noiseless and transferable. The approach enhances positive transfer from clean source examples to the target and mitigates negative transfer of noisy source examples. A thorough evaluation shows that our approach significantly outperforms the state-of-the-art on weakly-supervised domain adaptation 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
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Domain-Adversarial Training of Neural Networks
Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., Marchand, M., & Lempitsky, V. (2017) · Advances in Computer Vision and Pattern Recognition · doi:10.1007/978-3-319-58347-1_10 · Cited by 8,318
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 2×
“A rich literature has revealed that deep networks learn distributed representations that disentangle the explanatory factors of variations behind data which can reduce the domain discrepancy (Don-ahue et al. 2014; Yosinski et al. 2014).”
BackgroundBengio, Y., Louradour, J., Collobert, R., & Weston, J. (2009) · Proceedings of the 26th Annual International Conference on Machine Learning · doi:10.1145/1553374.1553380 · Cited by 5,366 · cited 3×
“This paper presents a Transferable Curriculum Learning ( TCL ) approach to weakly-supervised domain adaptation, extending from curriculum learning (Bengio et al. 2009) and adversarial learning (Goodfellow et al. 2014).”
BackgroundMethodAdversarial 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 10×
“In particular, TCL outperforms state-of-the-art deep domain adaptation methods DAN, DANN and ADDA with large margins since these methods suffer from negative transfer and over-fitting caused by noisy source examples.”
BackgroundInfluentialLearning 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 13×
“In particular, TCL outperforms state-of-the-art deep domain adaptation methods DAN, DANN and ADDA with large margins since these methods suffer from negative transfer and over-fitting caused by noisy source examples.”
BackgroundInfluentialSelf-citationDeCAF: 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 2×
“Feature Visualization: We visualize the t-SNE embeddings (Donahue et al. 2014) of the bottleneck representations by DAN, DANN, RTN and TCL on transfer task A →W.”
BackgroundMethod(2012) · MPG.PuRe (Max Planck Society) · doi:10.5555/2188385.2188410 · Cited by 3,603
Adapting 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×
“Office-31 (Saenko et al. 2010) is a standard dataset for domain adaptation, consisting of 4652 images with 31 classes in 3 distinct domains: Amazon (A), with images collected from amazon.”
BackgroundInfluentialUnderstanding deep learning requires rethinking generalization
Zhang, C., Bengio, S., Hardt, M., Recht, B., & Vinyals, O. (2016) · arXiv · doi:10.48550/arxiv.1611.03530 · Cited by 2,838 · cited 3×
“Thus, weakly-supervised domain adaptation is more general and challenging compared with standard domain adaptation, since the classification model tends to overfit on noisy source data (Zhang et al. 2017), degrading its generalization performance on the target domain.”
BackgroundInfluential
Cited by
A survey of unsupervised deep domain adaptation
Wilson, G., & Cook, D. J. (2020) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3400066 · Cited by 987
A Survey on Curriculum Learning
Wang, X., Chen, Y., & Zhu, W. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3069908 · Cited by 692
“(i) For inaccurate supervision, i.e., the training set is noisy and usually collected from the web, CL helps to denoise, enabling the model to focus on a cleaner subset to avoid bad local minimum [31], [64], [86], [103].”
BackgroundInfluentialSoviany, P., Ionescu, R. T., Rota, P., & Sebe, N. (2022) · International Journal of Computer Vision · doi:10.1007/s11263-022-01611-x · Cited by 431
A Review of Single-Source Deep Unsupervised Visual Domain Adaptation
Zhao, S., Yue, X., Zhang, S., Li, B., Zhao, H., Wu, B., Krishna, R., Gonzalez, J. E., Sangiovanni-Vincentelli, A. L., Seshia, S. A., et al. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2020.3028503 · Cited by 352
Memory-aware curriculum federated learning for breast cancer classification
Jiménez-Sánchez, A., Tardy, M., González Ballester, M. A., Mateus, D., & Piella, G. (2023) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2022.107318 · Cited by 114
“[35] addressed two entangled challenges of weakly-supervised DA: sample noise of the source domain, and distribution shift across domains.”
Background
Metadata
- Authors
- Shu, Y., Cao, Z., Long, M., & Wang, J.
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2019-07-17
- Volume
- 33
- Issue
- 01
- Pages
- 4951-4958
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W2904549000
- Cited by
- 105
- References
- 13
- FWCI
- 5.71
- Type
- conference paper
- Language
- en
- Open access
- diamond
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Transferable Curriculum for Weakly-Supervised Domain Adaptation. https://app.cochat.ai/papers/10.1609/aaai.v33i01.33014951
- The paper
- Shu, Y., Cao, Z., Long, M., & Wang, J. (2019). Transferable Curriculum for Weakly-Supervised Domain Adaptation. Proceedings of the AAAI Conference on Artificial Intelligence. https://doi.org/10.1609/aaai.v33i01.33014951
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.