Contrastive Representation Learning: A Framework and Review
Open accessThe DOI matches its Crossref record; 147 of the 162 references it cites were found in a registry or an index, and 15 are not checked yet.
Abstract
Contrastive Learning has recently received interest due to its success in self-supervised representation learning in the computer vision domain. However, the origins of Contrastive Learning date as far back as the 1990s and its development has spanned across many fields and domains including Metric Learning and natural language processing. In this paper, we provide a comprehensive literature review and we propose a general Contrastive Representation Learning framework that simplifies and unifies many different contrastive learning methods. We also provide a taxonomy for each of the components of contrastive learning in order to summarise it and distinguish it from other forms of machine learning. We then discuss the inductive biases which are present in any contrastive learning system and we analyse our framework under different views from various sub-fields of Machine Learning. Examples of how contrastive learning has been applied in computer vision, natural language processing, audio processing, and others, as well as in Reinforcement Learning are also presented. Finally, we discuss the challenges and some of the most promising future research directions ahead.
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 source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“According to the suggestions of Le-Khac et al. [29], the main methods can be divided into the following categories: multisensory signals, data augmentation, localglobal consistency, and temporal consistency.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [29]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
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 [56]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“As summarized in [29], commonly used contrastive loss functions generally include scoring functions (cosine similarity), energy-based margin functions (pair loss and triplet loss), probabilistic NCE-based functions, and mutual information based functions.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [29]
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 2×
“In SimCLRv2 [17], empirical experiments show that the output of the second layer of a 3-layer MLP projection head is a better representation for transfer learning than the output of the base ResNet [42] encoder.”
BackgroundMethodVaswani, 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×
“Recently, the Bidirectional Encoder Representation from Transformer (BERT) [24] model learns bidirectional word representations using the Transformer architecture’s decoder [103] and demonstrated great performance for transfer learning in multiple downstream tasks.”
MethodHochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 1×
“Even though the BERT [24] text encoder and the Faster RCNN [87] object detection model are trained separately and not updated by the contrastive loss, the LSTM cells [48] and a linear project head can still map between words and object representations.”
MethodExploiting 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
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
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 5×
“The discretised speech representation can then be used directly as a representation for other models that expect discrete input such as BERT [24].”
MethodInfluentialFaster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812
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
BERT: 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 5×
“The discretised speech representation can then be used directly as a representation for other models that expect discrete input such as BERT [24].”
MethodInfluentialEfficient Estimation of Word Representations in Vector Space
Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1301.3781 · Cited by 26,568
Research works printed but not linked: 1
- stratos, 2019, Noise contrastive estimation
Cited by
Items 601–700 of 948
Enhancing self-supervised visual representation learning through adversarially generated examples
Kang, M., & Kim, J. (2025) · Neural Computing and Applications · doi:10.1007/s00521-025-11236-z · Cited by 1
“This architecture eliminates the need for negative pairs by introducing a momentum encoder and relying solely on positive pairs for representation learning [17, 31].”
InfluentialEnhancing Intrusion Detection Systems with representation methods: A comparative study
Meyer, B. H., Pozo, A., Nogueira, M., & Zola, W. M. N. (2025) · doi:10.1109/cisdb64969.2025.11010407 · Cited by 1
“We will compare three well-known approaches [23]: Generative Adversarial Networks (GAN) [9], Autoencoders [26], and Contrastive Learning [13].”
Transfer learning-based few-shot sample deep learning-assisted template attacks
Feng, Z., Li, L., Ou, Y., & Deng, L. (2025) · Physical Communication · doi:10.1016/j.phycom.2025.102710 · Cited by 1
Balogh, O. M., Pétervári, M., Csernák, Á. M., Puhl, E., Horváth, A., Ferdinandy, P., & Ágg, B. (2026) · Briefings in Bioinformatics · doi:10.1093/bib/bbag463 · Cited by 1
Hu, Y., Wang, S., Hu, Q., & Li, J. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3576794 · Cited by 1
“At its core, this approach seeks to align semantically similar samples (positive samples) in the latent space while ensuring that dissimilar samples (negative samples) remain spaced apart [40].”
Convolutional attention-based contrastive learning for partial label learning
Gong, B., Zhou, R., & Jing, L. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06757-y · Cited by 1
“Le-Khac et al. [ 35] proposed a novel general framework for contrastive learning representations that simplifies and unifies various methods in this field.”
Efficient Compression of Mass Spectrometry Images via Contrastive Learning-Based Encoding
Radziński, P., Skrajny, J., Moczulski, M., Ciach, M. A., Valkenborg, D., Balluff, B., & Gambin, A. (2025) · Analytical Chemistry · doi:10.1021/acs.analchem.4c06913 · Cited by 1
Flight Maneuver Evaluation Method Based on Similarity Metric via Triplet Network
Wang, Q., Liu, C., Liu, S., Zhang, Y., & Yang, J. (2025) · IEEE Access · doi:10.1109/access.2025.3598248 · Cited by 1
“The fundamental objective of triplet networks is to learn an embedding space where semantically similar samples are clustered tightly, while dissimilar samples are separated by a margin [16].”
Zhu, R., Wang, Y., Shang, J., Dai, L.-Y., & Li, F. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.3098 · Cited by 2
“In recent years, contrastive learning ( Le-Khac, Healy & Smeaton, 2020 ), as an emerging learning method, has attracted great interest from a wide range of researchers.”
Learning a refinement model for variant analysis in non-human primate genomes
Choi, J., Zhou, B., & Song, G. (2025) · BMC Genomics · doi:10.1186/s12864-025-11921-2 · Cited by 1
“Recent studies have explored improving deep learning performance by refining input representations through contrastive learning [61] or optimizing attention mechanisms with additional contextual features [62].”
Zhang, Y., Liu, R., Chen, Q., Mo, K., Chu, L., & Kang, Z. (2025) · IEEE Access · doi:10.1109/access.2025.3602449 · Cited by 3
“…representation learning methods have empirically demonstrated significant effectiveness across robotic control and reinforcement learning domains [13], rigorous theoretical analyses addressing their application to highly uncertain, high-dimensional nonlinear systems in space remain relatively…”
P-Mix: A Data Augmentation Method for Contrastive Learning Based Human Activity Recognition
Chen, Y., Xie, Q., Cui, W., Chen, L., Song, H., & Zhu, T. (2025) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2025.3601599 · Cited by 1
Region-aware unknown information enhancement for open world object detection
Su, S., Wang, C., Zhu, Y., & Xu, Y. (2025) · Cluster Computing · doi:10.1007/s10586-025-05161-y · Cited by 1
“They introduced an energy function based on Faster R-CNN to distinguish whether the input sample belongs to an unknown class and incorporated contrastive learning [40] to further enhance the classifier’s performance.”
Contrastive Autoencoding with Gaussian Confidence Regions for Concept Drift Detection in IDS
Vieira, L., Choren, R., & Sant’ana, R. (2025) · Simpósio Brasileiro de Cibersegurança (SBSeg) · doi:10.5753/sbseg.2025.10493 · Cited by 1
“By enhancing feature discrimination, contrastive learning proves effective in representation learning, facial recognition, recommendation systems and anomaly detection [Le-Khac et al. 2020].”
Multimodal Data Dynamic Compression Algorithm Based on Semantic Importance
Wei, S., Feng, C., Guo, C., & Zhang, B. (2025) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops67674.2025.11162223 · Cited by 1
“To establish the association between multimodal data, We propose a Transformer-based semantic encoding architecture that establishes geometric alignment across modalities through contrastive learning [9].”
Tahtouh, M., Bernoux, G., Brunet, A., Standarovski, D., Nguyen, G., & Sicard, A. (2025) · Journal of Geophysical Research Machine Learning and Computation · doi:10.1029/2024jh000566 · Cited by 2
Self-Supervised Job Grouping for Parallel Machine Scheduling with Family Setups
Jeong, S., Kwon, C., & Kim, H.-J. (2025) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2025.09.172 · Cited by 1
FedSKC: Federated Learning with Non-IID Data via Structural Knowledge Collaboration
Wang, H., Li, H., Chen, H., Yan, J. H., Wang, L., Shi, J., Chen, S., & Shen, J. (2025) · IEEE International Conference on Web Services (ICWS) · doi:10.1109/icws67624.2025.00093 · Cited by 3
“Contrastive learning has been widely applied to computer vision in the self-supervised learning scenarios [25], [30]–[32].”
Hu, Y., Zhang, Q., Zhao, Z., Ding, X., Li, W., Yang, J., Sun, J., & Xu, L. X. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103844 · Cited by 4
Huang, H., Huang, Y., Du, X., Pang, J., Nong, S., Pan, X., Lan, M., Lu, L., Tan, Q., Gu, X., et al. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3625941 · Cited by 1
“The CSTCM model integrates the advantages of the swin transformer architecture [17], [18], [19], [20] and contrastive loss [21], [22], [23], [24], enhancing the model’s capability in detecting kidney stones, particularly under class-imbalanced and small-lesion conditions.”
Kim, Y., Lee, J., Lee, B. C., & Seo, H. (2025) · Medical Physics · doi:10.1002/mp.70078 · Cited by 1
“25”
Sun, S., Xia, X., & Hua, Z. (2025) · Journal of Intelligent Manufacturing · doi:10.1007/s10845-025-02711-4 · Cited by 1
“Contrastive learning is a self-supervised learning approach that learns discriminative high-dimensional representations by leveraging similarity or distance measures between data samples (Le-Khac et al., 2020 ).”
Invariance Pair Guidance: Robustness to Spurious Correlations via Corrective Gradients
Surner, M., Khelil, A., & Bothmann, L. (2026) · Machine Learning · doi:10.1007/s10994-026-07092-0 · Cited by 2
“The first three task gradients (blue) are scaled to two-thirds the length of the corrective gradient due to the violation of the invariance condition.”
Explainable ECG Analysis by Explicit Information Disentanglement With VAEs
van der Valk, V., Atsma, D. E., Scherptong, R. W. C., & Staring, M. (2025) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2025.3631143 · Cited by 1
“Thisdynamicfieldhaspro-ducedseveralprominentalgorithmsandmethods,includingbut notlimitedtoStyleGANs,VAEs,andcontrastivelearning[11], [12],[13],[14].”
Huang, Z. H., Zeng, Q., & Gai, N. (2025) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2025.103307 · Cited by 1
Hasegawa, K., Hidano, S., & Fukushima, K. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-4674-9_18 · Cited by 1
A Self-Supervised Contrastive Learning Approach for Detecting Deepfakes in Medical CT Scans
Pradeepan, P., Raj, S. G., George, J., & Nath, N. V. (2025) · Journal of Intelligent & Fuzzy Systems · doi:10.1177/18758967251398052 · Cited by 1
“Contrastive methods such as SimCLR and MoCo learn strong feature representations from unlabeled data by optimizing the feature similarity across different augmented views (Le-Khac et al., 2020).”
Physics-Guided Domain Alignment for Condition Monitoring of Three-Phase Inverters
Fassi, Y., Zhao, S., Ferretti, J., Wei, X., Heiries, V., Boutet, J., Boisseau, S., & Wang, H. (2025) · IEEE Energy Conversion Congress and Exposition (ECCE) · doi:10.1109/ecce58356.2025.11260182 · Cited by 1
Learning Context-aware Term Importance for Query Performance Prediction
Saleminezhad, A., Arabzadeh, N., Beheshti, S., & Bagheri, E. (2025) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3779428 · Cited by 1
“These evaluations demonstrate the effectiveness and stability of our approach compared to the state of the art.”
A multi-view cooperative contrastive learning framework for Birdsong Classification
Wang, Z., Yu, C., Zhang, Y., Lv, D., Cao, Y., & Zhao, Y. (2025) · Ecological Informatics · doi:10.1016/j.ecoinf.2025.103538 · Cited by 1
Margin-based angular losses for lightweight text classification: Lessons from face recognition
Jordehi, A. Y., & Mirroshandel, S. A. (2025) · Applied Soft Computing · doi:10.1016/j.asoc.2025.114431 · Cited by 1
CORAL: Disentangling Latent Representations in Long-Tailed Diffusion
Rodriguez, E., Welfert, M., McDowell, S., Stromberg, N., Camarena, J., & Sankar, L. (2025) · neural information processing systems · doi:10.52202/085713-3754 · Cited by 1
Sun, Y., Han, X., Yu, H., Du, Y., Guo, W., Li, X., Zhong, Y., Liu, Z., Zhang, D., & Bao, H. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3607884 · Cited by 1
“Therefore, inspired by self-supervised learning [12], [13], [14], [15], [16] and contrastive learning [17], [18], [19 Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
SSFormer: Self-supervised transformer model for predicting the properties of crystalline materials
Feng, J., Sun, T., & Yuan, J. (2025) · Computational Materials Science · doi:10.1016/j.commatsci.2025.114447 · Cited by 2
Zhang, G., Wang, C., Tao, Y., Zhou, H., Ma, Y., & Han, X. (2026) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3650277 · Cited by 1
“In contrast to conventional classification models, contrastive learning focuses more on intersample di ff erences and is particularly e ff ective with limited datasets [25].”
Unified Complementary Learning with Feature Perturbation for Semi-Supervised Learning
Wang, K., Yang, J., Guo, Y., & Xue, A. (2026) · Algorithms · doi:10.3390/a19010056 · Cited by 1
The Comparison of Human and Machine Performance in Object Recognition
Kul, G., & Wills, A. J. (2026) · Behavioral Sciences · doi:10.3390/bs16010109 · Cited by 1
Supervised Contrastive Frame Aggregation for Video Representation Learning
Chowdhury, S., Rahman, M., & Hamerly, G. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-14495-9_8 · Cited by 2
Kim, D. H., Bok, Y., Lee, S. H., Heo, J., Park, S., Lee, D., & Park, J. (2026) · Scientific Reports · doi:10.1038/s41598-025-34911-6 · Cited by 1
“The pre-training method of ECG-FM, initially proposed by Oh et al. 28 , is a lead-agnostic self-supervised approach that uses random lead masking and contrastive learning 29 , combining CNN, transformer architecture, and temporal average pooling to capture latent ECG representations (Fig.”
Zheng, W., Li, X., Cui, G., & Chen, S. (2026) · PLoS ONE · doi:10.1371/journal.pone.0341118 · Cited by 1
“Recent work has shown that domain-adapted contrastive objectives, which incorporate technical term relationships and hierarchical concept structures, outperform generic contrastive frameworks [40].”
SwCC: A Swapped-Contrastive Clustering Learning for Few-shot Website Fingerprinting Attacks
Bai, J., Gou, G., Gu, Z., Dong, W., Xiong, G., Li, Z., Li, Z., & Yang, Q. (2025) · IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) · doi:10.1109/trustcom66490.2025.00213 · Cited by 1
Physics-Informed Simulation-to-Hardware Alignement for DC-Link Capacitor Condition Monitoring
Fassi, Y., Zhao, S., Heiries, V., Boutet, J., Boisseau, S., & Wang, H. (2026) · IEEE Transactions on Power Electronics · doi:10.1109/tpel.2026.3663772 · Cited by 1
“This article introduces a unified framework that combines domain-aware contrastive learning [33], [34], [35] and adaptive loss balancing to monitor the health of DC-link capacitors using both synthetic and experimental data.”
SADGR: Adaptive Cross-Modal Emotion Recognition via Self-Supervised Alignment and Dynamic Gating
Zhang, J., Mao, J., Hou, Y., Noh, G., & Zhang, F. (2026) · IEEE Access · doi:10.1109/access.2026.3663574 · Cited by 1
“This method operates by using a meticulously designed objective function (like InfoNCE loss [17]) to minimize the distance between positive pairs (such as matched images and text) and maximize the distance between negative pairs (mismatched combinations) within a shared embedding space [18].”
scDCL: A multi-view single-cell RNA sequencing clustering method based on dual contrastive learning
Gan, L., Meng, H., Chen, Y., & Wang, Y. (2026) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2026.108998 · Cited by 1
Wei, X., Chen, J., Cai, F., & Shao, F. (2026) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2026.104768 · Cited by 1
Yang, C., Liang, Q., Guo, Z., Wang, F., & Liu, D. Y. (2026) · Information Processing in Agriculture · doi:10.1016/j.inpa.2026.03.007 · Cited by 1
Motion-Aware Representation Network for Micro-Expression Recognition
Xu, S., Chen, C. L. P., & Zhang, T. (2026) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2026.3675939 · Cited by 1
“The CNN-block downscales the concatenated features to obtain F motion . pairs [30].”
Industrial chain risk monitoring with explainable dynamic knowledge graph learning
Guo, Y., Ma, Y., Deng, S., Lv, Z., & Wang, X. (2026) · Array · doi:10.1016/j.array.2026.100796 · Cited by 1
GCRdetector: Graph contrastive representation learning for urban anomaly detection
Jiang, X., Shen, C., Yao, B., & Shao, S. (2026) · Information Sciences · doi:10.1016/j.ins.2026.123490 · Cited by 1
Li, G., Chen, N., Zhu, H., Niu, Y., Xu, Z., Li, J., & Zhu, Z. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116035 · Cited by 1
Towards evolution-based multi-level collaborative multi-task sparse learning
Liu, Y., Li, J., Gong, M., Zhang, Y., & Tang, Z. (2026) · Swarm and Evolutionary Computation · doi:10.1016/j.swevo.2026.102385 · Cited by 1
Hierarchical anchor-guided contrastive learning with heterogeneous GNN for multi-view clustering
Takrar, A., Tab, F. A., & Abdollahpouri, A. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116033 · Cited by 1
Wang, H., Chen, Z., Chen, K., Li, S., Liu, Y., Wang, P., & Yang, L. (2026) · Journal of Electroanalytical Chemistry · doi:10.1016/j.jelechem.2026.120221 · Cited by 1
Anyimadu, E. A., Zhang, X., Fuller, C. D., Marai, G. E., & Canahuate, G. M. (2026) · ACM Symposium on Applied Computing · doi:10.1145/3748522.3779983 · Cited by 1
“Originally developed for self-supervised learning, this paradigm learns latent embeddings by pulling similar samples closer together and pushing dissimilar ones apart within a latent space [16].”
Enhancing vehicle detection under adverse weather conditions with contrastive learning
Li, B., Liu, C., Kyösti, P., Öhman, M., Roy, D. S., Plazzi, S., Hagner, O., & Mokayed, H. (2026) · Image and Vision Computing · doi:10.1016/j.imavis.2026.106083 · Cited by 1
Zhu, H., Gao, W., Li, X., Qu, L., Li, W., & Hou, B. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3716678 · Cited by 1
“It mainly learns meaningful semantic feature representations from unlabeled data by encouraging similar samples to cluster together while separating dissimilar ones [23], [24].”
Contrastive Metric Learning for Point Cloud Segmentation in Highly Granular Detectors
Marriott-Clarke, M., Novakovic, L., Ratzer, E., Bainbridge, R., Gouskos, L., & Maier, B. (2026) · PRX Intelligence · doi:10.1103/tsc6-gn3r · Cited by 1
“Such objectives have been widely used in representation learning and clustering tasks [12,14,15], where downstream performance depends on the separability of learned features.”
Quantifying Representational Complexity in Transformer Models via Residual Stream Spectral Entropy
Barvaliya, K. (2026) · International Journal of Computer Applications · doi:10.5120/ijca54dc2360b436 · Cited by 3
Task-Free Online Replay with Contrastive Learning and Dynamic Herding
Biswas, R., Mohan, C. K., Nag, S., & Dandapat, S. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-31930-2_34 · Cited by 1
Self-Supervised Spatio-Temporal Contrastive Learning for FDIA Detection in the Smart Grid
Huang, X., & Zhang, Z. (2026) · IEEE/CAA Journal of Automatica Sinica · doi:10.1109/jas.2026.125741 · Cited by 1
Thailand Stock Similarity Clustering by Self-Supervised Wavelet transforms
Ploysuwan, T., & Pravithana, N. (2021) · International Conference on Big Data Analytics and Practices (IBDAP) · doi:10.1109/ibdap52511.2021.9552076 · Cited by 1
Chen, Z., Liang, Y., & Liu, Y. (2022) · 2022 7th International Conference on Signal and Image Processing (ICSIP) · doi:10.1109/icsip55141.2022.9885803 · Cited by 1
“Contrastive Learning revolves around learning a mapping from different views of the same scene, or context into the same region of representation space [19], which is formalised through the similarity distribution.”
BackgroundCLOP: Video-and-Language Pre-Training with Knowledge Regularizations
Li, G., Yang, H., He, F., Feng, Z., Lyu, Y., Wu, H. R., & Wang, H. (2022) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3503161.3548346 · Cited by 1
“The key idea behind contrastive learning [19] is to learn a general feature function by discriminating positive and negative samples – making the positive pairs attracted and the negative pairs separated [13, 42, 48].”
BackgroundContrastive Learning Based Visual Representation Enhancement for Multimodal Machine Translation
Wang, S., Zhang, W., Guo, W., Yu, D., & Liu, P. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892312 · Cited by 1
“The idea of contrastive learning is to pull positive pairs closer while pushing negative pairs further, thus the key of contrastive learning is how these pairs are generated [13], [27].”
BackgroundFault Diagnosis for Flight Control System Based on Contrastive Learning
An, Z., Cheng, Y., & Song, D. (2023) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-6613-2_342 · Cited by 1
Augmenting Feature Representation with Gradient Penalty for Robust Text Categorization
Wang, D., Cheng, L., & Wang, Z. (2023) · International Journal of Intelligent Systems · doi:10.1155/2023/7386888 · Cited by 1
Co-Loss Function for Rodents Sleep Stage Scoring Based on Single-Channel EEG
Chang, S., & You, Y. (2022) · China Automation Congress (CAC) · doi:10.1109/cac57257.2022.10055401 · Cited by 1
“Thus, it can overcome the limitation encountered in supervised learning settings [12].”
BackgroundContext-Dependent Deep Learning for Affective Computing
Suresh, V. (2022) · International Conference on Affective Computing and Intelligent Interaction Workshops and Demos (ACIIW) · doi:10.1109/aciiw57231.2022.10085993 · Cited by 1
“Many computer vision approaches use contrastive learning particularly in self-supervised settings [30]–[32].”
BackgroundEmploying Latent Categories of Entities for Knowledge Graph Embeddings With Contrastive Learning
Yang, J., Ying, X., & Chen, T. (2023) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2023.3268044 · Cited by 1
“Contrastive learning is an efficient representation learning method that contrasts positive pairs against negative pairs [26], [27].”
BackgroundWang, J., Tian, S., Feng, X., Zhang, B., Wu, F., Zhang, H., & Wang, C. (2023) · Remote Sensing · doi:10.3390/rs15143697 · Cited by 1
Sauron, T., Lazarus, C., Kurtz, C., & Cloppet, F. (2023) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi53787.2023.10230339 · Cited by 1
Universally Strict Black-Box Attacks Against Deep Neural Networks
Lu, Z., & Yang, Y. (2023) · International Conference on Image, Vision and Computing (ICIVC) · doi:10.1109/icivc58118.2023.10270265 · Cited by 1
Learning Pair-Centric Representation for Link Sign Prediction with Subgraph
Chen, J., Dai, F., Gu, X., Fan, H., Zhou, J., Li, B., & Wang, W. (2023) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3583780.3614951 · Cited by 4
“Recent works [15, 16, 45] have investigated the advantages of the balanced embedding space from contrastive learning [12, 22], whose main idea is to learn discriminative representations by contrasting positive and negative samples.”
BackgroundAsymmetric Graph Contrastive Learning
Chang, X., Wang, J., Guo, R., Wang, Y., & Li, W. (2023) · Mathematics · doi:10.3390/math11214505 · Cited by 1
Wang, J., Tian, S., Feng, X., Zhang, B., Wu, F., & Wang, C. (2023) · SAR in Big Data Era (BIGSARDATA) · doi:10.1109/bigsardata59007.2023.10294715 · Cited by 1
“We proposed a Dual Contrastive Learning Model for Representation Learning and Background Debiasing (RLDCL), where contrastive learning (CL) [9,10] is developed to draw semantic attention to object attributes and discourage the acquisition of background-learning features.”
MethodChannel Estimation Based on Contrastive Feature Learning with Few Labeled Samples
Xu, Y., & Lian, L. (2023) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc53906.2023.10304517 · Cited by 1
“Contrastive learning (CL) [11] has been used in the field of computer vision [12], natural language processing [13], image classification [14], etc, for feature extraction in self-supervised learning.”
BackgroundMultAtt-RippleNet: Multi-attribute and Knowledge-based Attention Fusion Recommendation Model
Chen, H., & Li, Q.-G. (2023) · Journal of Physics Conference Series · doi:10.1088/1742-6596/2644/1/012006 · Cited by 1
Meta-Learning and Self-Supervised Pretraining for Storm Event Imagery Translation
Rugina, I., Dangovski, R., Veillette, M., Khorrami, P., Cheung, B., Simek, O., & Marin Soljacic, A. (2023) · IEEE High Performance Extreme Computing Conference · doi:10.1109/hpec58863.2023.10363448 · Cited by 2
“Growing fields such as self-supervised [18] and multi-task learning [12] reflect these interests and provide promising solutions to the aforementioned issues.”
BackgroundChung, S.-C. (2023) · Journal of Structural Biology · doi:10.1016/j.jsb.2023.108058 · Cited by 1
Auto-Augmentation Contrastive Learning for Wearable-based Human Activity Recognition
Wu, Q., Shen, J., Fan, F., Gu, Y., Xu, C., & Chen, Y. (2023) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm58861.2023.10385468 · Cited by 3
“The contrastive learning framework relies on data augmentation strategies to construct different views of each sample to maximize pairwise agreement [1], [2].”
BackgroundDeep Self-supervised Subspace Clustering with Triple Loss
Bu, X., Dong, J., Zhang, M., Feng, G., Gao, X., & Niu, S. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-53308-2_31 · Cited by 1
Wei, R., Chen, H., Lyu, L., Zhang, Y., & Wang, J. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.116887 · Cited by 1
Shin, J., Kwak, J., & Jung, J. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e32929 · Cited by 3
“It is a method of learning a representation space by comparing input samples through self-supervision so that similar data are close and different data are far away in the representation space [46].”
MethodLingyun, B. A. O., HUANG, Z., Zehui, L. I. N., Sun, Y., Chen, H., Li, Y., Li, Z., Xiaochen, Y. U. A. N., Lin, X. U., & Tan, T. (2024) · Virtual Reality & Intelligent Hardware · doi:10.1016/j.vrih.2024.02.001 · Cited by 1
Silva, J. V. V., Barbosa, G. N. N., Lunardi, W. T., Lopez, M. A., & Mattos, D. M. F. (2024) · doi:10.5753/sbc.14597.4.2 · Cited by 1
DiagNCF: Diagnosis Neural Collaborative Filtering for Accurate Medical Recommendation
Pan, Q., & Zhang, J. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-5692-6_10 · Cited by 1
Reference-free review-based product question answering evaluation via distant contrastive learning
Huang, T. D., Ren, Y., Wu, L., & Zhang, X. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10649987 · Cited by 1
Pre-training Encoder-Decoder for Minority Language Speech Recognition
Wang, B., Xiaofeng, J., Yu, M., Wang, G., & Chen, J. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10649982 · Cited by 2
“Recently, the emergence of contrastive learning [7], relying on the comparison of acoustic features to deepen the understanding of mutual information between similar features, or using autoregressive reconstruction methods [8] to enhance feature capturing capabilities, has brought new possibilities…”
BackgroundGai, D., Wu, Y., Xiao, Y., Geng, Y., Cao, L., Xiong, X., & Zhong, A. (2024) · IET Image Processing · doi:10.1049/ipr2.13227 · Cited by 1
Nafea, M. S., & Ismail, Z. H. (2024) · IEEE Access · doi:10.1109/access.2024.3487532 · Cited by 2
“Recently, in unsupervised learning paradigms, the focus has greatly shifted towards representation learning [17].”
BackgroundAudio meets text: a loss-enhanced journey with manifold mixup and re-ranking
Suryawanshi, Y., Shah, V., Randar, S., & Joshi, A. D. (2024) · Knowledge and Information Systems · doi:10.1007/s10115-024-02283-4 · Cited by 2
“For this purpose, loss functions must define a scoring function and an optimization constraint [21].”
InfluentialNeuroMoCo: a neuromorphic momentum contrast learning method for spiking neural networks
Ma, Y., Wang, H., Shen, H., Chen, X., Duan, S., & Wen, S. (2024) · Applied Intelligence · doi:10.1007/s10489-024-05982-1 · Cited by 3
“This learning paradigm encompasses two primary branches: masked autoencoder[13] and contrastive learning[21].”
A Flat-Span Contrastive Learning Method for Nested Named Entity Recognition
Liu, Y., Zhang, K., Tong, R., Cai, C., Chen, D., & Wu, X. (2024) · International Journal of Asian Language Processing · doi:10.1142/s2717554524500139 · Cited by 1
Folco, M. D., Bernardino, G., Clarysse, P., & Duchateau, N. (2026) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2026.3698240 · Cited by 1
“It pursues objectives comparable to contrastive learning [15], but within a manifold learning perspective, namely providing a latent space into which distances have statistical meaning.”
Panda, S., Pareek, V., & Saxena, S. (2025) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.8379 · Cited by 1
“Initially, contrastive learning [37–39] was devised to address self-supervised tasks by enhancing data representations.”
Continual learning of conjugated visual representations through higher-order motion flows
Marullo, S., Tiezzi, M., Gori, M., & Melacci, S. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.107296 · Cited by 1
Embedding Spatial and Semantic Contexts for Geo-Entity Typing in Smart City Applications
Shbita, B., Vu, B., Lin, F., & Knoblock, C. A. (2025) · International World Wide Web Conference (WWW) · doi:10.1145/3701716.3718325 · Cited by 1
“Inspired by these advances, we apply self-supervised contrastive learning [12, 24] to geospatial representation learning, optimizing geo-entity embeddings for semantic typing and AI-driven urban analytics.”
A Graph Representation Learning‐Based Method for Event Prediction
Zeng, X., Luo, G., Qin, K., & Zheng, P. (2025) · IET Information Security · doi:10.1049/ise2/9706647 · Cited by 1
“Le-Khac Le-Khac et al. [17] reviewed graph comparison learning, highlighting its future potential.”
Textual and structural dual enhancement for knowledge graph completion with large language models
Wang, L., Gan, Y., Wang, X. A., Dong, Y., & Xu, Z. (2025) · Journal of Intelligent Information Systems · doi:10.1007/s10844-025-00953-5 · Cited by 1
Dfa-Con: A Contrastive Learning Approach for Detecting Copyright Infringement in Deepfake Art
Wahab, H., Ugail, H., & Mehmood, I. (2025) · Machine learning for signal processing ... · doi:10.1109/mlsp62443.2025.11204260 · Cited by 1
“Contrastive learning has shown great promise in learning discriminative and generalizable visual representations [8].”
Metadata
- Authors
- Le-Khac, P. H., Healy, G., & Smeaton, A. F.
- Affiliations
- Dublin City University (IE)
- Venue
- IEEE Access
- Published
- 2020
- Volume
- 8
- Pages
- 193907-193934
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W3096655658
- Cited by
- 862
- References
- 144
- Citations by year
- 2026: 111 · 2025: 197 · 2024: 212 · 2023: 157 · 2022: 113 · 2021: 71 · 2020: 1
- FWCI
- 33.61
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- IEEE Access (publishedVersion, cc-by)
- arXiv (Cornell University) (acceptedVersion)
- Dublin City University Open Access Institutional Repository (Dublin City University) (submittedVersion, cc-by)
- Arrow@dit (Dublin Institute of Technology) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- arXiv (Cornell University) (cc-by)
- Funders
- Science Foundation Ireland through the Science Foundation Ireland (SFI) Centre for Research Training in Machine Learning (18/CRT/6183)
- Insight Centre for Data Analytics (SFI/12/RC/2289_P2)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Algorithms
- Keywords
- contrastive learning, representation learning, self-supervised learning, metric learning, inductive bias, computer vision, natural language processing, audio signal processing, reinforcement learning
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.