Self-Supervised Representation Learning: Introduction, advances, and challenges
Open accessThe DOI matches its Crossref record; 110 of the 125 references it cites were found in a registry or an index, and 1 are not checked yet.
Abstract
Self-supervised representation learning (SSRL) methods aim to provide powerful, deep feature learning without the requirement of large annotated data sets, thus alleviating the annotation bottleneck—one of the main barriers to the practical deployment of deep learning today. These techniques have advanced rapidly in recent years, with their efficacy approaching and sometimes surpassing fully supervised pretraining alternatives across a variety of data modalities, including image, video, sound, text, and graphs. This article introduces this vibrant area, including key concepts, the four main families of approaches and associated state-of-the-art techniques, and how self-supervised methods are applied to diverse modalities of data. We further discuss practical considerations including workflows, representation transferability, and computational cost. Finally, we survey major open challenges in the field, that provide fertile ground for future work.
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.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“The various self-supervised pretexts are divided into five broad families: transformation prediction, masked prediction, instance discrimination, clustering, and contrastive instance discrimination.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [28]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Contrastive learning [56-58] is in fact a training procedure used in recent artificial intelligence models as a way to efficiently learn semantic representations.”
From How Adversarial REM Dreams May Facilitate Creativity, and Why We Become Aware of Them (2024) · cites this paper as [58]
The cited source says what the claim attributes to it.
In the citing paper
“For example, Ericsson et al. [28] provides a very comprehensive review of pretext tasks for multiple modalities, including image, video, text, audio, time series, and graph.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [28]
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 · cited 4×
“State of the art architectures usually start with CNN representation encoding hθ(·), with ResNet [1] being widely used, before appending task-specific”
InfluentialVaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · cited 3×
“long been used to model the context while recent works have moved to Transformer-based architectures with self-attention [2], which allow longer range connections to be made across words in a sentence, but require more data for training.”
Exploiting Generative AI to Scale up Intelligent Tutoring Systems
Jakubův, J., Chvalovský, K., Goertzel, Z., Kaliszyk, C., Olšák, M., Piotrowski, B., Schulz, S., Suda, M., & Urban, J. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.itp.2023.19 · Cited by 76,658
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173 · cited 9×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities, from image [5]–[7], video [8], [9], speech [10], [11], text [12], [13], and graphs [14], [15].”
InfluentialImageNet: 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 · cited 2×
“As in much of computer vision, ImageNet [35] is the most typical source data set for self-supervised pretraining [5]–[7], [40], consisting of 1.28 million training images across 1,000 object categories, with the most commonly used resolution at 224 × 224.”
InfluentialBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325 · cited 9×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities, from image [5]–[7], video [8], [9], speech [10], [11], text [12], [13], and graphs [14], [15].”
InfluentialAn Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2010.11929 · Cited by 47,472 · cited 1×
“One notable version is a vision transformer [58], which is increasingly being used by recent self-supervised methods on image data [43].”
The Elements of Statistical Learning
Hastie, T., Tibshirani, R., & Friedman, J. (2009) · Springer Series in Statistics · doi:10.1007/978-0-387-84858-7 · Cited by 25,970 · cited 2×
“These range from classic shallow approaches such as Gaussian mixtures [21] to deep methods such as variational autoencoders (VAEs) and generative adversarial networks (GANs) [18].”
Semi-supervised classification with graph convolutional networks
Kipf, T. N., & Welling, M. (2016) · arXiv · doi:10.48550/arxiv.1609.02907 · Cited by 24,853 · cited 1×
“Graph NNs (GNNs) [89] have advanced all these tasks significantly, especially where massive-labeled data are available.”
Distributed Representations of Words and Phrases and their Compositionality
Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1310.4546 · Cited by 24,533 · cited 6×
“With data annotation being a major bottleneck, NLP was the first discipline to make major and successful use of self-supervision [22].”
InfluentialLanguage Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · cited 9×
“Various data sets have been created from this data, and filtered versions of the entire corpus often form the bulk of training sets [32].”
InfluentialLearning Transferable Visual Models From Natural Language Supervision
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. (2021) · arXiv · doi:10.48550/arxiv.2103.00020 · Cited by 15,293 · cited 5×
“Other applications include speech [11]; and multi-view [41] and multi-modal representation learning including audiovisual [42] and visuo-linguistic [43] data – where matching and mismatching views of the same instance are contrasted against each other.”
InfluentialKaugeranna, R., Kaugeranna, E., & Aion, (. S. 4. 6. (2023) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.giscience.2023.43 · Cited by 13,690
Momentum Contrast for Unsupervised Visual Representation Learning
He, K., Fan, H., Wu, Y., Xie, S., & Girshick, R. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00975 · Cited by 13,347
The Elements of Statistical Learning: Data Mining, Inference, and Prediction
John Lu, Z. Q. (2010) · Journal of the Royal Statistical Society Series A: Statistics in Society · doi:10.1111/j.1467-985x.2010.00646_6.x · Cited by 12,390
Deep Contextualized Word Representations
Peters, M., Neumann, M., Iyyer, M., Gardner, M., Clark, C., Lee, K., & Zettlemoyer, L. (2018) · Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers) · doi:10.18653/v1/n18-1202 · Cited by 11,983 · cited 1×
“A final trend is that new models are becoming bigger, counting ELMo [56] at 94 M, BERT [12] at 340 M, GPT-2 [13] at 1.5 billion and GPT-3 [32] at 175 billion parameters.”
Influentialnode2vec: Scalable Feature Learning for Networks
Grover, A., & Leskovec, J. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939754 · Cited by 10,926 · cited 5×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities from image [5], [6], [7], video [8], [9], speech [10], [11], text [12], [13] and graphs [14], [15].”
InfluentialInductive Representation Learning on Large Graphs
Hamilton, W. L., Ying, R., & Leskovec, J. (2017) · arXiv · doi:10.48550/arxiv.1706.02216 · Cited by 10,639 · cited 3×
“Depending on whether the ultimate task of interest requires node-or graph-level predictions, methods may focus on learning node-[15], [27], [57] or graph-level [90] representations, or both [91].”
Lewis, M., Liu, Y., Goyal, N., Ghazvininejad, M., Mohamed, A., Levy, O., Stoyanov, V., & Zettlemoyer, L. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.703 · Cited by 8,821
Lewis, M., Liu, Y., Goyal, N., Ghazvininejad, M., Mohamed, A., Levy, O., Stoyanov, V., & Zettlemoyer, L. (2019) · arXiv · doi:10.48550/arxiv.1910.13461 · Cited by 8,816 · cited 3×
“Finally, although there may be a dominant pretext strategy for a given modality, it is common that suitably designed combinations of pretexts applied in a multitask manner can improve performance compared to a single pretext [81].”
InfluentialRepresentation Learning with Contrastive Predictive Coding
Oord, A. V. D., Li, Y., & Vinyals, O. (2018) · arXiv · doi:10.48550/arxiv.1807.03748 · Cited by 6,242 · cited 4×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities, from image [5]–[7], video [8], [9], speech [10], [11], text [12], [13], and graphs [14], [15].”
Influential“Librispeech: an ASR corpus based on public domain audio books,”
Panayotov, V., Chen, G., Povey, D., & Khudanpur, S. (2015) · 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2015.7178964 · Cited by 6,072 · cited 1×
“…for SSRL become baked into the resulting language models. reserved for downstream evaluation, with contemporary meth-ods being pretrained on Librispeech [67], containing 960 h of speech from audiobook readings, and Libri-light [68], a much larger data set (60,000 h of similar audiobook recordings).”
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
Baevski, A., Zhou, Y., Mohamed, A., & Auli, M. (2020) · neural information processing systems · Cited by 5,821
Context Encoders: Feature Learning by Inpainting
Pathak, D., Krahenbuhl, P., Donahue, J., Darrell, T., & Efros, A. A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.278 · Cited by 5,691 · cited 2×
“Common masking methods involve hiding words in sentences for language modeling [12], [22], [32], hiding time slices in speech [10], hiding regions of images for inpainting [26], or hiding edges in graphs [27].”
Dimensionality Reduction by Learning an Invariant Mapping
Hadsell, R., Chopra, S., & LeCun, Y. (2006) · 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2 (CVPR'06) · doi:10.1109/cvpr.2006.100 · Cited by 5,336 · cited 2×
“The earliest example of a self-supervised system, given the modern interpretation of the phrase, is the work of [37].”
Targeted Branching for the Maximum Independent Set Problem Using Graph Neural Networks
Silva, G., Rodrigues, M., Teixeira, A., & Amorim, M. (2024) · DROPS (Schloss Dagstuhl – Leibniz Center for Informatics) · doi:10.4230/lipics.sea.2024.20 · Cited by 5,183
Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning
Grill, J.-B., Strub, F., Altché, F., Tallec, C., Richemond, P. H., Buchatskaya, E., Doersch, C., Pires, B. A., Guo, Z. D., Azar, M. G., et al. (2020) · arXiv · doi:10.48550/arxiv.2006.07733 · Cited by 3,846 · cited 3×
“For images, the majority of recent work still uses ImageNet, with its 1.28 million images as the source set [6], [7].”
InfluentialAudio Set: An ontology and human-labeled dataset for audio events
Gemmeke, J. F., Ellis, D. P. W., Freedman, D., Jansen, A., Lawrence, W., Moore, R. C., Plakal, M., & Ritter, M. (2017) · 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2017.7952261 · Cited by 3,134 · cited 1×
“An additional data set considered here is AudioSet [69], an audio event detection data set.”
MoleculeNet: a benchmark for molecular machine learning
Wu, Z., Ramsundar, B., Feinberg, E. N., Gomes, J., Geniesse, C., Pappu, A. S., Leswing, K., & Pande, V. (2018) · Chemical Science · doi:10.1039/c7sc02664a · Cited by 3,107 · cited 2×
“In such applications data are intrinsically hard to collect, but predicting graph properties can significantly impact tasks such as drug discovery and material discovery [92], [93].”
Noise-contrastive estimation: A new estimation principle for unnormalized statistical models
Gutmann, M. U., & Hyvärinen, A. (2010) · Edinburgh Research Explorer (University of Edinburgh) · Cited by 2,992
ImageNet: A large-scale hierarchical image database
Deng, J. H., Dong, W. F., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvprw.2009.5206848 · Cited by 2,714
Deep Clustering for Unsupervised Learning of Visual Features
Caron, M., Bojanowski, P., Joulin, A., & Douze, M. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01264-9_9 · Cited by 2,674
A survey on semi-supervised learning
van Engelen, J. E., & Hoos, H. H. (2020) · Machine Learning · doi:10.1007/s10994-019-05855-6 · Cited by 2,666 · cited 2×
“This setting can also be addressed by semi-supervised meth-ods [20], which should then be evaluated as competitors against SSRL. ■ If the target domain of interest is similar enough to large source data sets (e.g., everyday objects versus ImageNet), then one can leverage self-supervised pretraining…”
The Kinetics Human Action Video Dataset
Kay, W., Carreira, J., Simonyan, K., Zhang, B., Hillier, C., Vijayanarasimhan, S., Viola, F., Green, T., Back, T., Natsev, P., et al. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1705.06950 · Cited by 2,593 · cited 6×
“In video and multi-modal settings, common transfer evaluation considers transfer from large source datasets such as Kinetics [61] to standard target datasets such as UCF101.”
InfluentialUnsupervised Learning of Visual Features by Contrasting Cluster Assignments
Caron, M., Misra, I., Mairal, J., Goyal, P., Bojanowski, P., & Joulin, A. (2020) · arXiv · doi:10.48550/arxiv.2006.09882 · Cited by 2,499 · cited 10×
“As in much of computer vision, ImageNet [35] is the most typical source data set for self-supervised pretraining [5]–[7], [40], consisting of 1.28 million training images across 1,000 object categories, with the most commonly used resolution at 224 × 224.”
InfluentialRevisiting Unreasonable Effectiveness of Data in Deep Learning Era
Sun, C., Shrivastava, A., Singh, S., & Gupta, A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.97 · Cited by 2,484 · cited 1×
“However, the conventional paradigm has been to train these systems using supervised learning, where performance has grown roughly logarithmically with annotated data set sizes [3].”
Unsupervised Representation Learning by Predicting Image Rotations
Gidaris, S., Singh, P., & Komodakis, N. (2018) · arXiv · doi:10.48550/arxiv.1803.07728 · Cited by 2,477 · cited 3×
“; in which case, we might want to use a TP technique like that which is detailed in [33].”
Influentialwav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations
Baevski, A., Zhou, H., Mohamed, A., & Auli, M. (2020) · arXiv · doi:10.48550/arxiv.2006.11477 · Cited by 2,369 · cited 6×
“A notable success in speech was shown by wav2vec 2.0 [10], which used transformers plus masked prediction SSRL on 53,000 h of unlabeled data prior to fine-tuning a down-stream speech recognition system.”
InfluentialDeep Speech 2: End-to-End Speech Recognition in English and Mandarin
Amodei, D., Anubhai, R., Battenberg, E., Case, C. T., Casper, J., Catanzaro, B., Chen, J., Chrzanowski, M., Coates, A., Diamos, G., et al. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.02595 · Cited by 2,110 · cited 1×
“Meanwhile, contemporary NN approaches trained by supervised learning have dominated in settings where massive, annotated training data are available [83].”
ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks
Lu, J., Batra, D., Parikh, D., & Lee, S. (2019) · arXiv · doi:10.48550/arxiv.1908.02265 · Cited by 2,109 · cited 1×
“Notably, ViLBERT [55] exploited both BERT-like masked prediction and contrastive-instance discrimination to learn a multimodal representation, which then achieved state-of-the-art performance in downstream vision and language tasks such as caption-based retrieval, visual question answering (QA),…”
Self-Supervised Visual Feature Learning With Deep Neural Networks: A Survey
Jing, L., & Tian, Y. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.2992393 · Cited by 2,036 · cited 2×
“Compared to existing surveys [4], we provide a broader introduction to the field; a wider coverage of different modalities rather than focusing on images; highlight more practical considerations such as representation transferability, computation cost, and deployment strategies; and provide a…”
Tian, Y., Krishnan, D., & Isola, P. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58621-8_45 · Cited by 1,997
TinyBERT: Distilling BERT for natural language understanding
Jiao, X., Yin, Y., Shang, L., Jiang, X., Chen, X., Li, L., Wang, F., & Liu, Q. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.372 · Cited by 1,786
Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020) · Communications of the ACM · doi:10.1145/3381831 · Cited by 1,647
The kinetics human action video dataset
Kay, W. (2024) · TIB Data Manager · doi:10.57702/jtuxnt5m · Cited by 1,646
Cross-lingual Language Model Pretraining
Lample, G., & Conneau, A. (2019) · arXiv · doi:10.48550/arxiv.1901.07291 · Cited by 1,586 · cited 3×
“One can pretrain SSRL models on large corpora of high-resource languages before fine-tuning them on smaller corpora of low-resource languages [25].”
YFCC100M: the new data in multimedia research
Thomee, B., Shamma, D. A., Friedland, G., Elizalde, B., Ni, K., Poland, D., Borth, D., & Li, L.-J. (2016) · Communications of the ACM · doi:10.1145/2812802 · Cited by 1,541
The iNaturalist Species Classification and Detection Dataset
Van Horn, G., Mac Aodha, O., Song, Y., Cui, Y., Sun, C., Shepard, A., Adam, H., Perona, P., & Belongie, S. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00914 · Cited by 1,451 · cited 1×
“…with certain biases that do not appear in natural images, such as centering of objects and clear isolation of object from background. iNaturalist [60] is a collection of wildlife data sets compiled by a citizen science project, where members upload their own photographs and others collectively…”
Exploring the Limits of Weakly Supervised Pretraining
Mahajan, D., Girshick, R., Ramanathan, V., He, K., Paluri, M., Li, Y., Bharambe, A., & van der Maaten, L. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01216-8_12 · Cited by 1,224
Justifying Recommendations using Distantly-Labeled Reviews and Fine-Grained Aspects
Ni, J., Li, J., & McAuley, J. (2019) · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) · doi:10.18653/v1/d19-1018 · Cited by 1,200 · cited 3×
“Like in other modalities we have seen increasingly large graphs being used for pre-training, like the Amazon Review Recommendation data [71] with 113 million nodes or Open Academic Graph (OAG) [70] consists of over 178 million nodes and 2 billion edges.”
Protein function prediction via graph kernels
Borgwardt, K. M., Ong, C. S., Schonauer, S., Vishwanathan, S. V. N., Smola, A. J., & Kriegel, H.-P. (2005) · Bioinformatics · doi:10.1093/bioinformatics/bti1007 · Cited by 1,105 · cited 1×
“G ra ph Open Academic Graph [70] 178M nodes, 2B edges Amazon Review Recommendation [71] 113M nodes PROTEINS [72] 1.”
SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems
Wang, A., Pruksachatkun, Y., Nangia, N., Singh, A., Michael, J., Hill, F., Levy, O., & Bowman, S. R. (2019) · arXiv · doi:10.48550/arxiv.1905.00537 · Cited by 1,051 · cited 3×
“This has been done successfully in NLP and has driven progress, making sure that it benefits many areas of the field [80], but such standardized benchmarks are lacking in the other modalities we have considered.”
InfluentialSun, F.-Y., Hoffmann, J., Verma, V., & Tang, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1908.01000 · Cited by 1,027 · cited 3×
“Several recent techniques optimize mutual information-based, instance discrimination objectives, with DGI [15] and InfoGraph [90] performing contrastive-instance discrimination between pairs of nodes/patches and whole graphs.”
(2020) · International Conference on Machine Learning · Cited by 1,012
HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
Miech, A., Zhukov, D., Alayrac, J.-B., Tapaswi, M., Laptev, I., & Sivic, J. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00272 · Cited by 986 · cited 2×
“For video pretraining, state-of-the-art models use the increasingly large YouTube-8M-2 [62] and HowTo100M [63] with combined video playtimes of 13 and 15 years, respectively.”
Learning Representations for Automatic Colorization
Larsson, G., Maire, M., & Shakhnarovich, G. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_35 · Cited by 946 · cited 1×
“The others mask out information in the training images and require the network to reconstruct it, leading to pretext tasks such as colorization [53] and inpainting [26], where color channels and image patches, respectively, are removed.”
What makes for good views for contrastive learning
Tian, Y., Sun, C., Poole, B., Krishnan, D., Schmid, C., & Isola, P. (2020) · arXiv · doi:10.48550/arxiv.2005.10243 · Cited by 773 · cited 1×
“Recent work has systematically demonstrated this intuition that the ideal transformations to use do indeed depend on the downstream task [44].”
Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks
Dosovitskiy, A., Fischer, P., Springenberg, J. T., Riedmiller, M., & Brox, T. (2016) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2015.2496141 · Cited by 762
Barlow Twins: Self-Supervised Learning via Redundancy Reduction
Žbontar, J., Jing, L., Misra, I., LeCun, Y., & Deny, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.03230 · Cited by 751 · cited 1×
“There are many different schemes, like using asymmetrical encoding for the two inputs [6] or minimizing redundancy via cross correlation between features [39].”
Audio-Visual Scene Analysis with Self-Supervised Multisensory Features
Owens, A., & Efros, A. A. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01231-1_39 · Cited by 704
Revisiting Self-Supervised Visual Representation Learning
Kolesnikov, A., Zhai, X., & Beyer, L. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00202 · Cited by 675 · cited 1×
“There have been mixed results reported in the literature with regard to whether linear readout is sufficient or whether fine-tuning the entire encoder should improve performance [29], [30].”
Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty
Hendrycks, D., Mazeika, M., Kadavath, S., & Song, D. (2019) · arXiv · doi:10.48550/arxiv.1906.12340 · Cited by 634 · cited 4×
“4) Furthermore, SSRL leads to better calibrated probabilities [28], [29], which can be used to drive abstention of automated predictions or out-of-distribu-tion detection [28].”
InfluentialSelf-supervised learning for medical image analysis using image context restoration
Chen, L., Bentley, P., Mori, K., Misawa, K., Fujiwara, M., & Rueckert, D. (2019) · Medical Image Analysis · doi:10.1016/j.media.2019.101539 · Cited by 575 · cited 1×
“For example, the authors in [75] used unlabeled brain scan images to perform image restoration (an inpainting-like task), improving upon random initialization for fine-tuning several downstream tasks.”
Debiased Contrastive LearningPreprint
Chuang, C.-Y., Robinson, J., Lin, Y.-C., Torralba, A. B., & Jegelka, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.00224 · Cited by 534 · cited 1×
“When this assumption is violated, they suffer from false-positive pretext task labels [45].”
Predicting multicellular function through multi-layer tissue networks
Zitnik, M., & Leskovec, J. (2017) · Bioinformatics · doi:10.1093/bioinformatics/btx252 · Cited by 519 · cited 4×
“The current state of the graph modality is that transfer-ability is good to unseen nodes within the same graph and to unseen graphs within the same data set, e.g., protein-protein interactions [93].”
InfluentialLibri-Light: A Benchmark for ASR with Limited or No Supervision
Kahn, J., Riviere, M., Zheng, W., Kharitonov, E., Xu, Q., Mazare, P. E., Karadayi, J., Liptchinsky, V., Collobert, R., Fuegen, C., et al. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9052942 · Cited by 458 · cited 1×
“…for SSRL become baked into the resulting language models. reserved for downstream evaluation, with contemporary meth-ods being pretrained on Librispeech [67], containing 960 h of speech from audiobook readings, and Libri-light [68], a much larger data set (60,000 h of similar audiobook recordings).”
(2021) · International Conference on Learning Representations · Cited by 449
GPT-GNN: Generative Pre-Training of Graph Neural Networks
Hu, Z., Dong, Y., Wang, K., Chang, K.-W., & Sun, Y. (2020) · Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3394486.3403237 · Cited by 446 · cited 6×
“A minority of methods are inductive [27], [57] in that they learn embedding functions that do not depend on a specific choice of input graph, and thus can be transferred to new target nodes or graphs.”
InfluentialLearning Correspondence From the Cycle-Consistency of Time
Wang, X., Jabri, A., & Efros, A. A. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00267 · Cited by 445 · cited 1×
“Nonetheless, there has been an uptake of SSRL methods in applications such as tracking [96] and detection.”
Pointer Sentinel Mixture Models
Merity, S., Xiong, C., Bradbury, J., & Socher, R. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1609.07843 · Cited by 427 · cited 1×
“The early word embeddings made heavy use of Wikipedia articles [64], or crawls of news sites and social media sites like Twitter.”
Big Self-Supervised Models are Strong Semi-Supervised Learners
Chen, T., Kornblith, S., Swersky, K., Norouzi, M., & Hinton, G. E. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.10029 · Cited by 421 · cited 10×
“As in much of computer vision, ImageNet [35] is the most typical source data set for self-supervised pretraining [5]–[7], [40], consisting of 1.28 million training images across 1,000 object categories, with the most commonly used resolution at 224 × 224.”
InfluentialSelf-Supervised Spatiotemporal Learning via Video Clip Order Prediction
Xu, D., Xiao, J., Zhao, Z., Shao, J., Xie, D., & Zhuang, Y. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.01058 · Cited by 415 · cited 3×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities, from image [5]–[7], video [8], [9], speech [10], [11], text [12], [13], and graphs [14], [15].”
Deep Anomaly Detection Using Geometric Transformations
Golan, I., & El‐Yaniv, R. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.10917 · Cited by 372 · cited 1×
“For example, current state-of-the-art anomaly detectors [77] rely on SSRL training of rotation prediction, with the rotation prediction accuracy providing the anomaly score.”
Rethinking Pre-training and Self-training
Zoph, B., Ghiasi, G., Lin, T.-Y., Cui, Y., Liu, H., Cubuk, E. D., & Le, Q. V. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.06882 · Cited by 369 · cited 1×
“A preliminary result in computer vision suggests not [99].”
An Unsupervised Autoregressive Model for Speech Representation Learning
Chung, Y.-A., Hsu, W.-N., Tang, H., & Glass, J. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-1473 · Cited by 357
Anomaly Detection in Video via Self-Supervised and Multi-Task Learning
Georgescu, M.-I., Barbalau, A., Ionescu, R. T., Shahbaz Khan, F., Popescu, M., & Shah, M. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.01255 · Cited by 350 · cited 2×
“Across these modalities, it can also be applied to boost diverse downstream tasks, including not only simple recognition but also detection and localization [16], dense prediction (signal transformation) [16], anomaly detection [17], and so on.”
How Well Do Self-Supervised Models Transfer?
Ericsson, L., Gouk, H., & Hospedales, T. M. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.00537 · Cited by 341 · cited 9×
“There have been mixed results reported in the literature with regard to whether linear readout is sufficient or whether fine-tuning the entire encoder should improve performance [29], [30].”
InfluentialSelf-citationDeep Transformer Models for Time Series Forecasting: The Influenza Prevalence Case
Wu, N., Green, B., Ben, X., & O’Banion, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2001.08317 · Cited by 335 · cited 1×
“In terms of time-series forecasting, transformers based on sequential masked prediction pretext have significantly out-performed conventional autoregressive models in predicting disease transmission [88].”
Scaling and Benchmarking Self-Supervised Visual Representation Learning
Goyal, P., Mahajan, D., Gupta, A., & Misra, I. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00649 · Cited by 330 · cited 7×
“Furthermore, some results suggest that self-supervised representation quality is also a logarithmic function of the amount of unlabeled pretraining data [16].”
InfluentialUnsupervised Scalable Representation Learning for Multivariate Time Series
Franceschi, J.-Y., Dieuleveut, A., & Jaggi, M. (2019) · arXiv · doi:10.48550/arxiv.1901.10738 · Cited by 305 · cited 1×
“In terms of representation learning on more general time-series data, masked prediction methods based on a trans-former architecture have been shown to match supervised state of the art in a suite of benchmarks in diverse application areas [86].”
Clustering Stability: An Overview
Ulrike, V. L. (2010) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000008 · Cited by 301 · cited 1×
“An example of how further work could address gaps in our current understanding is to extend theoretical frameworks analyzing (shallow) clustering methods [51] to the deep SSRL paradigm.”
Self-Supervised Learning by Cross-Modal Audio-Video Clustering
Alwassel, H., Mahajan, D., Korbar, B., Torresani, L., Ghanem, B., & Tran, D. (2019) · arXiv · doi:10.48550/arxiv.1911.12667 · Cited by 298 · cited 5×
“Self-supervised strategies have been leveraged successfully to improve sample efficiency of learning across a variety of modalities, from image [5]–[7], video [8], [9], speech [10], [11], text [12], [13], and graphs [14], [15].”
InfluentialSelf-supervised Co-training for Video Representation Learning
Han, T., Xie, W., & Zisserman, A. (2020) · arXiv · doi:10.48550/arxiv.2010.09709 · Cited by 292 · cited 3×
“State-of-the-art procedures successfully leverage large source data sets and approaches but do not yet outperform supervised pretraining [78].”
Veličković, P., Fedus, W., Hamilton, W. L., Lió, P., Bengio, Y., & Hjelm, R. D. (2018) · Apollo (University of Cambridge) · doi:10.17863/cam.40744 · Cited by 279
Benchmarking Representation Learning for Natural World Image Collections
Horn, G. V., Cole, E., Beery, S., Wilber, M. J., Belongie, S., & Aodha, O. M. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.01269 · Cited by 270
Learning Problem-Agnostic Speech Representations from Multiple Self-Supervised Tasks
Pascual, S., Ravanelli, M., Serrà, J., Bonafonte, A., & Bengio, Y. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2605 · Cited by 219
Self-Supervised Learning by Cross-Modal Audio-Video Clustering
Alwassel, H., Mahajan, D. K., Korbar, B., Torresani, L., Ghanem, B., & Tran, D. (2020) · King Abdullah University of Science and Technology Repository (King Abdullah University of Science and Technology) · Cited by 216
Online Deep Clustering for Unsupervised Representation Learning
Zhan, X., Xie, J., Liu, Z., Ong, Y.-S., & Loy, C. C. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00672 · Cited by 203
Self-citationA Theoretical Analysis of Contrastive Unsupervised Representation Learning
Arora, S., Khandeparkar, H., Khodak, M., Plevrakis, O., & Saunshi, N. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.09229 · Cited by 191 · cited 2×
“…approaches to connecting SSRL with conventional statistical learning theory: one approach that applies to only instance discrimination methods [48], another that primarily considers how SSRL learns useful representations for natural language tasks [49], and finally, a paper that makes use of…”
Self-Supervised MultiModal Versatile Networks
Alayrac, J.-B., Recasens, A., Schneider, R. G., Arandjelović, R., Ramapuram, J., Fauw, J. D., Smaira, L., Dieleman, S., & Zisserman, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.16228 · Cited by 189 · cited 3×
“…several studies [41], [79] have explored how SSRL on multimodal source data can be used to learn a stronger representation for single-modality downstream tasks, and ultimately outperform single-modality pretraining on diverse downstream tasks in unimodal video, still-image, or audio domains [79].”
Unsupervised Pretraining Transfers Well Across Languages
Riviere, M., Joulin, A., Mazare, P.-E., & Dupoux, E. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9054548 · Cited by 186 · cited 1×
“However, very rapid progress is currently being made in multilingual [97] speech models and cross-lingual transfer [98], so prospects for transferability seem promising.”
Strategies for Pre-training Graph Neural Networks
Hu, W., Liu, B., Gomes, J. S., Žitnik, M., Liang, P., Pande, V. S., & Leskovec, J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.12265 · Cited by 158 · cited 1×
“Depending on whether the ultimate task of interest requires node-or graph-level predictions, methods may focus on learning node-[15], [27], [57] or graph-level [90] representations, or both [91].”
Large-Scale Multilingual Speech Recognition with a Streaming End-to-End Model
Kannan, A., Datta, A., Sainath, T. N., Weinstein, E., Ramabhadran, B., Wu, Y., Bapna, A., Chen, Z., & Lee, S. (2019) · Interspeech 2019 · doi:10.21437/interspeech.2019-2858 · Cited by 155 · cited 1×
“However, very rapid progress is currently being made in multilingual [97] speech models and cross-lingual transfer [98], so prospects for transferability seem promising.”
Video Cloze Procedure for Self-Supervised Spatio-Temporal Learning
Luo, D., Liu, C., Zhou, Y., Yang, D., Ma, C., Ye, Q., & Wang, W. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i07.6840 · Cited by 147
Self-Supervised Learning for ECG-Based Emotion Recognition
Sarkar, P., & Etemad, A. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9053985 · Cited by 146 · cited 3×
“In temporal data, such as videos and other time series, one can shuf-fle the temporal order of signal samples and force the network to predict the original order [8], [34].”
N-gram Counts and Language Models from the Common Crawl
Buck, C., Heafield, K., & van Ooyen, B. (2014) · Language Resources and Evaluation Conference (LREC) · doi:10.63317/3uyduafd26my · Cited by 137 · cited 1×
“Another example is Common Crawl [66], a nonprofit project that makes data from billions of web pages freely accessible.”
Reducing Sentiment Bias in Language Models via Counterfactual Evaluation
Huang, P.-S., Zhang, H., Jiang, R., Stanforth, R., Welbl, J., Rae, J., Maini, V., Yogatama, D., & Kohli, P. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.7 · Cited by 124
Demystifying Contrastive Self-Supervised Learning: Invariances, Augmentations and Dataset Biases
Purushwalkam, S., & Gupta, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2007.13916 · Cited by 122 · cited 2×
“If this was replaced with scene images of multiple cluttered objects, then typical instance discrimination tasks like mapping two different crops of one image to the same identity could create false-positive pretext label noise that maps different semantic objects to the same representation [95].”
3D Self-Supervised Methods for Medical Imaging
Taleb, A., Loetzsch, W., Danz, N., Severin, J., Gaertner, T., Bergner, B., & Lippert, C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.03829 · Cited by 119 · cited 1×
“This has also recently inspired various extensions of standard pretext tasks into 3D [76].”
Tao, C., Qi, J., Lu, W., Wang, H., & Li, H. (2022) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2020.3038420 · Cited by 115 · cited 1×
“Beyond common benchmarks, SSRL has been successfully applied in application areas where labeled data are sparse, such as Earth-observation remote sensing [74].”
Self-training Improves Pre-training for Natural Language Understanding
Du, J., Grave, E., Gunel, B., Chaudhary, V., Celebi, O., Auli, M., Stoyanov, V., & Conneau, A. (2021) · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2021.naacl-main.426 · Cited by 106
OAG: Toward Linking Large-scale Heterogeneous Entity Graphs
Zhang, F., Liu, X., Tang, J., Dong, Y., Yao, P., Zhang, J., Gu, X., Wang, Y., Shao, B., Li, R., et al. (2019) · Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3292500.3330785 · Cited by 97 · cited 2×
“Like in other modalities, we have seen increasingly large graphs being used for pretraining, like the Amazon Review Recommendation data [71] with 113 million nodes or the Open Academic Graph [70], which consists of more than 178 million nodes and 2 billion edges.”
GPT-GNN: Generative Pre-Training of Graph Neural NetworksPreprint
Hu, Z., Dong, Y., Wang, K., Chang, K.-W., & Sun, Y. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.15437 · Cited by 84 · cited 6×
“A minority of methods are inductive [27], [57] in that they learn embedding functions that do not depend on a specific choice of input graph, and thus can be transferred to new target nodes or graphs.”
InfluentialSubject-Aware Contrastive Learning for Biosignals
Cheng, J. Y., Goh, H., Dogrusoz, K., Tuzel, O., & Azemi, E. (2020) · arXiv · doi:10.48550/arxiv.2007.04871 · Cited by 74 · cited 2×
“For example, using TP SSRL prior to training ECG-based emotion recognition [34] and contrastive-instance discrimination SSRL prior to learning downstream, EEG-based, motor-movement classification and ECG-based anomaly detection [87].”
YFCC100M: The New Data in Multimedia Research
Thomee, B., Shamma, D. A., Friedland, G., Elizalde, B., Ni, K., Poland, D., Borth, D., & Li, L.-J. (2015) · arXiv · doi:10.48550/arxiv.1503.01817 · Cited by 69
Gao, X., Hu, W., & Qi, G.-J. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00719 · Cited by 60
Predicting What You Already Know Helps: Provable Self-Supervised Learning
Lee, J. D., Lei, Q., Saunshi, N., & Zhuo, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.01064 · Cited by 45 · cited 2×
“It is shown in [50] that if there is conditional independence between x i and z i given the downstream label (and optionally some additional latent variables), then any model that successfully predicts z i from x s must be estimating the label (and optional latent variables).”
Distance-Based Regularisation of Deep Networks for Fine-TuningPreprint
Gouk, H., Hospedales, T. M., & Pontil, M. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.08253 · Cited by 41 · cited 2×
“Of course, there are numerous ways to con-trol the amount of fine-tuning allowed in terms of learning rate, and explicitly regularizing the fine-tuning step to prevent it from overfitting by limiting the deviation from the initial pretrained conditions [19].”
Self-citationA Mathematical Exploration of Why Language Models Help Solve Downstream Tasks
Saunshi, N., Malladi, S., & Arora, S. (2020) · arXiv · doi:10.48550/arxiv.2010.03648 · Cited by 32 · cited 2×
“…to only instance discrimination methods [48], another that primarily considers how SSRL learns useful representations for natural language tasks [49], and finally, a paper that makes use of conditional independence to further elucidate how masked prediction pretext tasks lead to useful…”
A Mathematical Exploration of Why Language Models Help Solve Downstream TasksPreprint
Saunshi, N., Malladi, S., & Arora, S. (2021) · arXiv (Cornell University) · Cited by 13
Chen, T., Wang, Z., Li, L., Li, D., Li, Z., Chang, X., Bian, P., Liang, G., Wang, Q., & Xie, T. (2025) · Schloss Dagstuhl – Leibniz-Zentrum für Informatik · doi:10.4230/lipics.ecoop.2025.6
Research works printed but not linked: 4
- 0, Youtube-8m A large and diverse labeled video dataset for video understanding research
- gokaslan, 2019, OpenWebText Corpus
- radford, 2019, Language Models are Unsupervised Multitask Learners
- goodfellow, 2016, Deep Learning
Cited by
SPot-the-Difference Self-supervised Pre-training for Anomaly Detection and Segmentation
Zou, Y., Jeong, J., Pemula, L., Zhang, D., & Dabeer, O. (2022) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-20056-4_23 · Cited by 651
BockNet: Blind-Block Reconstruction Network With a Guard Window for Hyperspectral Anomaly Detection
Wang, D., Zhuang, L., Gao, L., Sun, X., Huang, M., & Plaza, A. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3335484 · Cited by 579
“Self-SL comprises two key steps: 1) constructing input samples and corresponding supervision information on unlabeled data and 2) designing the supporting pretext task according to specific target tasks, aiming to extract the intrinsic data features that are profitable to the target task [59].”
BackgroundZhao, Z., Alzubaidi, L., Zhang, J., Duan, Y., & Gu, Y. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122807 · Cited by 410
A review of deep learning techniques for speech processing
Mehrish, A., Majumder, N., Bharadwaj, R., Mihalcea, R., & Poria, S. (2023) · Information Fusion · doi:10.1016/j.inffus.2023.101869 · Cited by 389
Self-supervised speech representation learning: A review
Mohamed, A., Lee, H.-Y., Borgholt, L., Havtorn, J. D., Edin, J., Igel, C., Kirchhoff, K., Li, S.-W., Livescu, K., Maaløe, L., et al. (2022) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2022.3207050 · Cited by 372
“Due to the popularity of SSL, reviews have been published about the technology in general [16], [17], [18] as well as its application to natural language processing (NLP) [19], [20],”
Background
Metadata
- Authors
- Ericsson, L., Gouk, H., Loy, C. C., & Hospedales, T. M.
- Affiliations
- University of Edinburgh (GB)
- Nanyang Technological University (SG)
- Venue
- IEEE Signal Processing Magazine
- Published
- 2022-05
- Volume
- 39
- Issue
- 3
- Pages
- 42-62
- ISSN
- 1053-5888 (ISSN-L)
- 1558-0792
- OpenAlex
- W3207924272
- Cited by
- 433
- References
- 125
- Citations by year
- 2026: 89 · 2025: 105 · 2024: 132 · 2023: 86 · 2022: 20
- FWCI
- 28.22
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Signal Processing Magazine (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by-nd)
- DR-NTU (Nanyang Technological University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Engineering and Physical Sciences Research Council (EP/S000631/1, EP/L016427/1, EP/R026173/1)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Algorithms
- Keywords
- self-supervised representation learning, pre-training, deep feature learning, image representation, video representation, sound representation, text representation, graph representation, computational cost, open problems
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.