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 501–600 of 948
Prototype-Augmented Contrastive Learning for Few-Shot Unsupervised Domain Adaptation
Gong, L., Zhang, W., Li, M., Zhang, J., & Zhang, Z. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-40292-0_17 · Cited by 3
Janda, A., Wagstaff, B., Ng, E. G., & Kelly, J. (2023) · Canadian Conference on Computer and Robot Vision · doi:10.1109/crv60082.2023.00026 · Cited by 5
“Following the formulation in [31], a query point xi ∈ R is first sampled from the dataset.”
MethodModel-Agnostic Bias Measurement in Link Prediction
Schwertmann, L., Ravi, M. P. K., & de Melo, G. (2023) · Findings of the Association for Computational Linguistics: EACL · doi:10.18653/v1/2023.findings-eacl.121 · Cited by 3
“SimKGC uses the InfoNCE loss with an additive margin (Le-Khac et al., 2020).”
MethodContrastive Learning for Reducing False Negatives with Global and local views in Augmented Data
Auh, J., Cho, C.-S., & Kim, S.-T. (2023) · Innovations in Intelligent Systems and Applications Conference (ASYU) · doi:10.1109/asyu58738.2023.10296635 · Cited by 3
“The self-supervised learning showed remarkable performance in various areas such as computer vision [6] and natural language processing [7], [8], [9] by learning meaningful representations from unlabeled data.”
BackgroundLiu, X., Su, Y., Wu, N., Ji, Y., Ji, R., & Lu, M. (2023) · Communications in computer and information science · doi:10.1007/978-981-99-8132-8_25 · Cited by 3
Contrastive Subspace Clustering with Dissimilarity Regularization
Chen, Y., Wu, W., & Ou-Yang, L. (2023) · International Conference on Machine Learning and Cybernetics · doi:10.1109/icmlc58545.2023.10327943 · Cited by 3
“Recently, contrastive learning [14, 15] has realized significant advances in unsupervised representation learning by pulling features of positive pairs close while pushing away those of negative pairs.”
BackgroundTest-Time Training for Deep MS/MS Spectrum Prediction Improves Peptide Identification
Ye, J., He, X., Wang, S., Dong, M., Wu, F., Lu, S., & Feng, F. (2023) · Journal of Proteome Research · doi:10.1021/acs.jproteome.3c00229 · Cited by 3
Cross-Model Comparative Loss for Enhancing Neuronal Utility in Language Understanding
Zhu, Y., Pang, L., Wu, K., Lan, Y., Shen, H., & Cheng, X. (2024) · ACM Transactions on Information Systems · doi:10.1145/3652599 · Cited by 4
Lian, Y., Shi, Y., Liu, Z., Jiang, B., & Li, X. (2024) · IET Image Processing · doi:10.1049/ipr2.13142 · Cited by 3
Data-driven definition and modelling of plant growth
Wichitwechkarn, V., Rohde, W., Fox, C., & Choudhary, R. (2024) · Smart Agricultural Technology · doi:10.1016/j.atech.2024.100495 · Cited by 3
Deep contrastive multi-view clustering with doubly enhanced commonality
Yang, Z., Zhu, C., & Li, Z. (2024) · Multimedia Systems · doi:10.1007/s00530-024-01400-1 · Cited by 3
“In this framework, algorithms are trained to maximize the similarity between similar data points while minimizing the similarity between dissimilar ones [34].”
Contrastive Unsupervised Representation Learning With Optimize-Selected Training Samples
Cheng, Y., Zhang, Z., Li, X., & Wang, S. (2024) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2024.3424331 · Cited by 3
“Subsequently, representations of sentences and images have been employed by [6], [7], [12], [15], and [16] respectively.”
Han, H., Lian, C., Xu, B., Zeng, Z., Alhudhaif, A., & Polat, K. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.112372 · Cited by 3
Domain Adaptation for Semantic Segmentation of Autonomous Driving with Contrastive Learning
Li, Q., Yang, S., Xu, M., Ren, B., Yan, X., & Zhou, F. (2024) · IEEE International Conference on Industrial Informatics (INDIN) · doi:10.1109/indin58382.2024.10774392 · Cited by 3
Cano, P., Musulén, E., & Gil, D. (2025) · International Journal of Computer Assisted Radiology and Surgery · doi:10.1007/s11548-024-03313-w · Cited by 3
“Self-supervised methods [11] allow learning representation spaces that capture abstract structures useful for performing downstream tasks.”
Enhancing Multi-agent Coordination via Dual-channel Consensus
Zhang, Q., Wang, K., Ruan, J., Yang, Y., Xing, D., & Xu, B. (2024) · Machine Intelligence Research · doi:10.1007/s11633-023-1464-2 · Cited by 3
Building robust traffic classifier under low quality data: A federated contrastive learning approach
Qin, T., Cheng, G., Yin, Z., Wei, Y., Yao, Z., & Chen, Z. (2025) · Digital Communications and Networks · doi:10.1016/j.dcan.2025.06.001 · Cited by 4
Topology only pre-training: towards generalised multi-domain graph models
Davies, A., Green, R., Filho, T. M. S., & Ajmeri, N. (2026) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-026-01210-1 · Cited by 5
“As pre-training, then fine-tuning on downstream tasks, representation learning leads to improved performance (Chen et al 2020b; Le-Khac et al 2020; Devlin et al 2018).”
A Sensorimotor Perspective on Contrastive Multiview Visual Representation Learning
Laflaquière, A. (2021) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2021.3086267 · Cited by 5
“Our goal here is not to give an exhaustive overview of this field (see, for instance, [31] and [32] for a review).”
BackgroundSelf-supervised Pretraining for Covid-19 and Other Pneumonia Detection from Chest X-ray Images
Hao, Y., Wang, Y., & Wang, X. (2022) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-030-89698-0_102 · Cited by 3
Triplet Contrastive Learning for Brain Tumor ClassificationPreprint
Liu, T. Y., & Feng, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.03611 · Cited by 3
“Contrastive Learning Contrastive learning is an effective technique for selfsupervised learning [14].”
MethodMeta-Learning and Self-Supervised Pretraining for Real World Image TranslationPreprint
Rugina, I., Dangovski, R., Veillette, M., Khorrami, P., Cheung, B., Simek, O., & Marin Soljacic, A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2112.11929 · Cited by 3
“Growing fields such as self-supervised (Le-Khac et al., 2020) or multi-task learning (Hospedales et al.”
BackgroundSelf-supervised learning for early detection of neurodegenerative diseases with small data
Jiang, H. (2023) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/166402 · Cited by 9
Enhancing Peak Assignment in 13C NMR Spectroscopy: A Novel Approach Using Multimodal AlignmentPreprint
Xu, H., Zhou, Z., & Hong, P. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2311.13817 · Cited by 4
Khan, H., Zaidi, S. F. A., Shah, P. M., Balakrishnan, K., Khan, R., Waqas, M., & Wu, J. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.00311 · Cited by 3
“However, euclidean distance has been shown to produce less robust and generalizable features compared to contrastive learning frameworks [92].”
Estimating the Empowerment of Language Model AgentsPreprint
Song, J., Gore, J., & Kleiman-Weiner, M. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.22504 · Cited by 4
“Successor Representation and Mutual Information Objective: We combine state and action embeddings by simple addition to obtain the joint representation used in the InfoNCE loss.”
InfluentialA Foundation Model for Electrodermal Activity DataPreprint
Alchieri, L., Garzon, M., Alecci, L., Bona, F. B. D., Gjoreski, M., Felice, G. D., & Silvia Santini, B. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2603.16878 · Cited by 3
“The latent embeddings often contain information which allows them to be used effectively in downstream tasks [56].”
BidirLM: From Text to Omnimodal Bidirectional Encoders by Adapting and Composing Causal LLMsPreprint
Boizard, N., Deschamps-Berger, T., Gisserot-Boukhlef, H., Hudelot, C., & Colombo, P. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2604.02045 · Cited by 7
“Consequently, the second paradigm, which is now dominant in practice, skips the masking phase entirely and applies contrastive learning directly (Le-Khac et al., 2020; Wang et al., 2024a; Lee et al., 2025).”
Qin, Y., Chu, R., & Liu, B. (2024) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001424550139 · Cited by 2
SAR-EdgeYOLO: Robust Bridge Detection in Low-Resolution SAR via Edge Enhancement
Han, S., Song, J., Kim, H., & Kim, D. (2024) · IEEE Access · doi:10.1109/access.2024.3480049 · Cited by 2
“This objective is achieved using a softmax function over the cosine similarity between pairs of representations [38].”
MethodAutoML-Guided Fusion of Entity and LLM-Based Representations for Document Classification
Koloski, B., Pollak, S., Navigli, R., & Škrlj, B. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-78977-9_7 · Cited by 3
PC-Match: Semi-Supervised Learning With Progressive Contrastive and Consistency Regularization
Kang, M., Seo, S., & Min, M. (2025) · IEEE Access · doi:10.1109/access.2025.3549465 · Cited by 2
“This approach [34], [35], [36] compares data distance based on similarity.”
Li, M., Wang, Y., Ni, Y., Xiong, H., Mei, Z., & Zhang, W. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3554520 · Cited by 2
“It has shown remarkable performance in various fields, such as natural language processing [28].”
Wang, S., Ge, Y., Liu, Q., & Zeng, Y. (2025) · Multimedia Systems · doi:10.1007/s00530-025-01824-3 · Cited by 3
Rieger, W. J., Bodén, M., Arnold, F. H., & Mora, A. (2025) · eLife · doi:10.1101/2025.06.13.659624 · Cited by 2
“Data engineering, such as choosing a dataset to enhance specific relationships 30 , enables domain expertise to inform pair schemes for training.”
Zhang, L., Ding, R. F., Chen, X., Hou, J., Si, D., Wang, Y., Lin, K., & Cao, R. (2026) · Protein Science · doi:10.1002/pro.70734 · Cited by 2
MOSS: Multi-Modal Representation Learning on Sequential Circuits
Wang, M., Sun, B., Mu, J., Gu, F., Han, B., Yang, T., Zhang, X., Liu, S., Wen, Y., Wang, H., et al. (2025) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac63849.2025.11132437 · Cited by 4
“These approaches leverage different modalities’ strengths to enhance the performance of representation learning [31].”
Mortlock, T., Chen, L., Smereka, J. M., Khargonekar, P. P., & Faruque, M. A. A. (2025) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2025.3612873 · Cited by 3
FluentEditor2: Text-Based Speech Editing by Modeling Multi-Scale Acoustic and Prosody Consistency
Liu, R., Xi, J., Jiang, Z., & Li, H. (2025) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslpro.2025.3624913 · Cited by 3
“2) Contrastive Global Prosody Consistency Constraint ( L CGPC ), which employs a contrastive learning mechanism [26] to ensure that the prosodic features of the edited region remain consistent with the surrounding context of the original utterance while distinguishing it from other utterances in…”
Cui, Z., Sun, X., Xing, R., Song, K., & Chen, X. (2025) · International Journal of Fatigue · doi:10.1016/j.ijfatigue.2025.109439 · Cited by 2
Yuan, H., Jing, X., Yan, Y., & Luo, W. (2025) · Advanced Intelligent Systems · doi:10.1002/aisy.202500818 · Cited by 2
Learning physical interactions to compose biological large language models
Clark, J. D., Dean, T. J., & Shukla, D. (2026) · Communications Chemistry · doi:10.1038/s42004-025-01883-7 · Cited by 2
“Contrastive learning encompasses a variety of related methods originally proposed for representation learning in computer vision 20 .”
scMFE: A multi-view fusion enhanced graph contrastive learning method for scRNA-seq data clustering
Bai, Y., Shu, Z., Tan, K., Zhang, Y., & Yu, Z. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.132662 · Cited by 2
Zhang, C., Torres, R. D. S., Gansel, L. C., & Bracke, M. B. M. (2026) · Smart Agricultural Technology · doi:10.1016/j.atech.2026.101814 · Cited by 3
Visual Analytics using Tensor Unified Linear Comparative Analysis
Okami, N., Miyake, K., Sakamoto, N., Nonaka, J., & Fujiwara, T. (2026) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2025.3633912 · Cited by 2
“1 There is another ML scheme also called contrastive learning, which aims to learn an embedding from similar and dissimilar instance pairs to place similar instances close together in the embedding space [36].”
Kabir, M. S., Kurkin, S., Paunova, R., Stoyanov, D., & Hramov, A. E. (2026) · Chaos Solitons & Fractals · doi:10.1016/j.chaos.2026.118234 · Cited by 2
Jiang, Y., Miao, Q., Dewei, W. U., He, J., & Zhao, C. (2026) · Drones · doi:10.3390/drones10050339 · Cited by 2
Contrastive learning of graph encoder for accelerating pedestrian trajectory prediction training
Yao, Z., Yu, J., & Ding, J. (2021) · IET Image Processing · doi:10.1049/ipr2.12185 · Cited by 2
“Contrastive unsupervised learning is the most significant method in the unsupervised learning area, and the encoding of the object is suitable to employ contrastive learning [12].”
MethodProb-CLR: A probabilistic approach to learn discriminative representation
Xie, X., Zhu, M., Wang, Y., & Zhang, L. (2021) · Knowledge-Based Systems · doi:10.1016/j.knosys.2021.107329 · Cited by 3
Stock Volatility Forecast Base on Comparative Learning and Autoencoder Framework
Du, Y., Li, Q., Zhang, Z., & Liu, Y. (2022) · doi:10.1145/3523111.3523126 · Cited by 2
“Contrastive learning [5] is widely used in unsupervised representation learning in the image domain.”
BackgroundSupervised Contrastive Meta-learning for Few-Shot Classification
Zhang, C., Wang, J., Huang, Z., Kong, L., Qu, X., Cheng, N., & Xiao, J. Z. (2022) · IEEE International Conference on High Performance Computing and Communications; IEEE International Conference on Smart City; IEEE International Conference on Data Science and Systems (HPCC/SmartCity/DSS) · doi:10.1109/hpcc-dss-smartcity-dependsys57074.2022.00262 · Cited by 2
“The invariant properties of the learned representation depend on how the pairs are formed [24, 25].”
BackgroundResearch on Semantic Segmentation Method of Remote Sensing Image Based on Self-supervised Learning
Zhang, W., & Wang, A. (2023) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2023.0140855 · Cited by 3
“Contrastive self-supervised learning [24], also referred to as contrastive learning, shows more promising results in the field of remote sensing compared to generative self-supervised learning.”
BackgroundLane Change Detection with an Ensemble of Image-based and Video-based Deep Learning Models
Nalçakan, Y., & Baştanlar, Y. (2023) · Innovations in Intelligent Systems and Applications Conference (ASYU) · doi:10.1109/asyu58738.2023.10296634 · Cited by 2
“Self-supervised contrastive representation learning has recently gained popularity due to its achievements in computer vision [16], [17].”
BackgroundAMGC is a multiple-task graph neutral network for epigenetic target profiling
Gu, S., Bao, L., Yang, Y., Zhao, Y., Tong, H. H. Y., Liu, L., Liu, H., Hou, T., & Kang, Y. (2024) · Cell Reports Physical Science · doi:10.1016/j.xcrp.2024.101850 · Cited by 2
Contrastive Learning Framework for Bitcoin Crash Prediction
Liu, Z., Shu, M., & Zhu, W. (2024) · Stats · doi:10.3390/stats7020025 · Cited by 2
Genre Classification of Books in Russian with Stylometric Features: A Case Study
Vanetik, N., Tyamanova, M., Kogan, G., & Litvak, M. (2024) · Information · doi:10.3390/info15060340 · Cited by 4
“As a representation learning technique, contrastive learning seeks to maximize similarity between samples in the same class and minimize it across samples in different classes [26].”
BackgroundMethodA self-supervised fusion for infrared and visible images via multi-level contrastive auto-encoding
Su, H., Nie, R., Cao, J., Zhang, Y., & Huang, J. (2024) · Infrared Physics & Technology · doi:10.1016/j.infrared.2024.105421 · Cited by 2
Multiple representation contrastive self-supervised learning for pulmonary nodule detection✰
Torki, A., Adibi, P., & Kashani, H. B. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112307 · Cited by 2
Pretraining Enhanced RNN Transducer
Lu, J., Lian, R., Jiang, D., Song, Y., Su, Z., Wei, V. J., & Yang, L. (2024) · CAAI Artificial Intelligence Research · doi:10.26599/air.2024.9150039 · Cited by 2
Improving Unsupervised Hierarchical Representation With Reinforcement Learning
An, R., Li, Y., He, X., Gu, P., Zhao, M., Dong, L., Hao, J., Wang, C., An, B., & Zhou, M. (2024) · 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/cvpr52733.2024.02165 · Cited by 2
“…the model to learn low-level information shared on different datasets and high-level semantic information in a hierarchical manner. proaches [47, 79], which utilize a deep generative model [21, 31, 40, 43, 68, 81] to capture the posterior distribution of the underlying explanatory factors…”
BackgroundKapetadimitri, G., & Hristu‐Varsakelis, D. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72344-5_8 · Cited by 2
Enhancing hyperspectral image prediction with contrastive learning in low-label regimes
Haidar, S., & Jose Oramas, M. (2026) · Applied Intelligence · doi:10.1007/s10489-025-07071-3 · Cited by 2
“Among the various approaches within the SSL, Contrastive Learning (CL) [10, 11] has emerged as an up-and-coming method.”
LLM-Augmented Symbolic RL with Landmark-Based Task Decomposition
Kheirandish, A., Xu, D., & Fekri, F. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10889431 · Cited by 2
“For this purpose, we have used contrastive learning [7] with the logic-predicate representation of the states as its input.”
Ren, S., Wang, H., Wang, K., Liu, R., & Wang, Z. (2025) · IEEE Access · doi:10.1109/access.2025.3551604 · Cited by 2
“The Siamese Neural Network [12] is a metric learning network based on contrastive learning [13].”
Coughprint: Distilled Cough Representations From Speech Foundation Model Embeddings
Laska, B., Xi, P., Valdés, J. J., Wallace, B., Green, J. R., Goubran, R. A., & Knoefel, F. D. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3568985 · Cited by 2
“Neural network-based representation learning aims to train a network to map input data into an abstract feature vector form that is useful for downstream tasks, without the need for hand-crafted or engineered features [20].”
Ersoy, S., Ersoy, E. O., Danış, A., & Türkoğlu, Ş. A. (2025) · Cerebral Cortex · doi:10.1093/cercor/bhaf148 · Cited by 3
A deep learning approach for objective evaluation of microscopic neuro-drilling craniotomy skills
Kumar, R., Dhanakshirur, R. R., Singh, R., Suri, A., Kalra, P. K., & Arora, C. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110650 · Cited by 2
Zero-shot object detection based on cross-modal guided clustering
Cheng, D., Xu, X., Zhang, H., Song, T., Jiang, H., & Kou, Q. (2025) · Image and Vision Computing · doi:10.1016/j.imavis.2025.105664 · Cited by 2
Wu, Y., Xiao, L., & Du, X. (2025) · Network · doi:10.3390/network5040045 · Cited by 2
UWBKey: Using Contrastive Learning for Efficient Secure Key Generation in UWB
Chen, H., & Dhekne, A. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3770630 · Cited by 2
“…before quantizing into bits, where the latent features exhibit better reciprocity and randomness than the original CIR. Contrastive learning [35] is a machine learning technique where a model learns representations by contrasting similar (positive) and dissimilar (negative) data samples.”
A Framework Using Contrastive Learning for Classification with Noisy Labels
Ciortan, M., Dupuis, R., & Peel, T. (2021) · Data · doi:10.3390/data6060061 · Cited by 13
Towards Accurate DGA Detection based on Siamese Network with Insufficient Training Samples
Hu, X., Li, M., Cheng, G., Li, R., Wu, H., & Gong, J. W. (2022) · International Conference on Communications · doi:10.1109/icc45855.2022.9838409 · Cited by 2
“It is pointed out that contrastive learning is a solution for small-scale data samples [15], [16].”
Dynamic Bayesian Contrastive Predictive Coding Model for Personalized Product Search
Wu, B., Meng, Z., & Liang, S. (2023) · ACM Transactions on the Web · doi:10.1145/3609225 · Cited by 3
“Contrastive learning [36] is one of the most popular methods used to enhance the representation learning and improve the efficient use of data, generality, and robustness of the model, which encourages embeddings to be closer between positive pairs and apart between negative pairs.”
MethodSpeaker Recognition Based on the Joint Loss Function
Feng, T., Fan, H., Ge, F., Cao, S., & Liang, C. (2023) · Electronics · doi:10.3390/electronics12163447 · Cited by 2
SCRAPS: Speech Contrastive Representations of Acoustic and Phonetic Spaces
Vallés-Ṕerez, I., Beringer, G., Biliński, P., Cook, G., & Barra-Chicote, R. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230540 · Cited by 3
Contrastive Alignment for Deep Multi-view Subspace Clustering
Wang, J., Wu, B., Zhang, H., & Zhou, Y. (2023) · International Conference on Pattern Recognition and Artificial Intelligence (PRAI) · doi:10.1109/prai59366.2023.10331958 · Cited by 2
“Thus by contrastive learning, the similarities of positive pairs will be maximized while that of negative pairs will be minimized [14].”
BackgroundSemi-supervised News Discourse Profiling with Contrastive Learning
Li, M., & Huang, R. (2023) · International Joint Conference on Natural Language Processing and Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (IJCNLP-AACL) · doi:10.18653/v1/2023.ijcnlp-main.49 · Cited by 3
Industry 4.0 Approach in Intelligent Manufacturing: Adoptability of Machine Learning and Challenges
Sharma, A., & Gupta, P. (2024) · Advances in logistics, operations, and management science book series · doi:10.4018/979-8-3693-1363-3.ch004 · Cited by 2
Multi-Reference Generative Face Video Compression with Contrastive Learning
Konuko, G., & Valenzise, G. (2024) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp61759.2024.10743797 · Cited by 3
“Contrastive learning [8], [12] is a widely recognized framework for representation learning, notable for its effectiveness in self-supervised learning tasks.”
BackgroundGeomCA: Geometric Evaluation of Data RepresentationsPreprint
Poklukar, P., Varava, A., & Kragić, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.12486 · Cited by 8
“…for example disentangled (Pfau et al., 2020; Kim & Mnih, 2018), in the form of a specific manifold or curvature (Arvanitidis et al., 2018; 2020; Moor et al., 2020; Schönenberger et al., 2020), or with particular similarity as learned by contrastive learning algorithms (Le-Khac et al., 2020).”
BackgroundGraph Symbiosis Learning.Preprint
Zeng, L., Xu, J. A., Yao, Z., Zhu, Y., & Li, J. (2021) · arXiv (Cornell University) · Cited by 2
Feature diversity in self-supervised learningPreprint
Malviya, P., & Sudhakar, A. V. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2209.01275 · Cited by 2
Hashemi, H. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2403.13825 · Cited by 7
Improving Node Representation by Boosting Target-Aware Contrastive LossPreprint
Lin, Y.-C., & Neville, J. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.03901 · Cited by 2
Rieger, W. J., Bodén, M., Arnold, F. H., & Mora, A. (2025) · eLife · doi:10.7554/elife.108186.1 · Cited by 4
Contrastive Diffusion Guidance for Spatial Inverse ProblemsPreprint
Basu, S., Amballa, C., Xu, Z., Sampedro, J. V., Nelakuditi, S., & Choudhury, R. R. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2509.26489 · Cited by 3
“To achieve this, we train f φ and g ψ contrastively using an InfoNCE-style loss function Le-Khac et al. (2020); Oord et al. (2018).”
Simeone, O. (2022) · Cambridge University Press eBooks · doi:10.1017/9781009072205.010 · Cited by 1
Borek-Marciniec, W., & Ksieniewicz, P. (2023) · Applied Intelligence · doi:10.1007/s10489-023-04878-w · Cited by 1
EEG based Parkinson Detection through Supervised Information Enhanced Contrastive Learning
Song, J., Li, X., Jiang, W., Wang, C., Zhao, Z., Lv, J., & Hu, B. (2023) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm58861.2023.10385841 · Cited by 1
“Contrastive Learning is a self-supervised learning algorithm [18].”
BackgroundA Contrastive Learning Method for Multi-Label Predictors on Hyperspectral Images
Haidar, S., & Jose Oramas, M. (2023) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers61460.2023.10430726 · Cited by 1
“Contrastive learning (CL) has a long standing history [3, 4, 5], but its recent application in self-supervised representation learning for computer vision [6] has drawn significant attention.”
BackgroundIoT Network Intrusion Detection using Contrastive Learning with a Lightweight Autoencoder
Zhang, X., Ngai, E. C. ‐., & Yang, S. (2023) · IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10449070 · Cited by 1
“Contrastive learning [7], [8] is a promising approach to alleviate false positives by improving the IDS’s ability to distinguish between benign and malicious traffic.”
BackgroundZhong, S., Guo, S., Liu, J., Huang, H., Wang, L., Li, J., Li, C., & Hei, Y. (2024) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3645098 · Cited by 1
“Contrastive learning aims to learn a low-dimensional embedding space where similar samples are grouped closely together while dissimilar ones are pushed further apart [18].”
BackgroundUnified Contextualized Knowledge Embedding Method for Static and Temporal Knowledge Graph
Gao, Y., Qiao, L., Huang, Z., Kan, Z., He, Y., & Li, D. (2024) · IEEE Transactions on Audio Speech and Language Processing · doi:10.1109/taslp.2024.3507557 · Cited by 1
“Contrastive Learning has gained prominence as a self-supervised learning technique, involving the comparison of positive and negative samples [58].”
Rasmussen, T. E., Sørensen, S., Pisinger, D., Jørgensen, T. M., & Baum, A. (2024) · Computers & Operations Research · doi:10.1016/j.cor.2024.106960 · Cited by 1
Data-free Knowledge Distillation based on GNN for Node Classification
Zeng, X., Liu, T., Zeng, M., Wu, Q., & Wang, M. (2025) · Lecture notes in computer science · doi:10.1007/978-981-97-5779-4_16 · Cited by 1
Neuro-Symbolic Contrastive Learning for Cross-domain Inference
Liu, M., Ueda, R., Wan, Z., Inoue, K., & Willcocks, C. G. (2025) · Electronic Proceedings in Theoretical Computer Science · doi:10.4204/eptcs.416.6 · Cited by 2
“Hard examples, or those data pairs that are challenging for the model to correctly classify due to their nuanced differences or similarities, are particularly crucial in the training process of contrastive learning [26, 36].”
Enhancing ASD Diagnosis with Contrastive and Non-Contrastive Models from Neuroimaging Data
Jha, A., Mehta, I. S., Khan, K., & Katarya, R. (2024) · International Conference on Mobile Networks and Wireless Communications (ICMNWC) · doi:10.1109/icmnwc63764.2024.10872187 · Cited by 1
“The contrastive learning model in this study leverages the SimCLR (Simple Contrastive Learning of Representations) framework, which maximizes agreement between augmented views of the same rs-fMRI scan while minimizing agreement between views from different scans [5].”
Fusion of Time-Frequency Features in Contrastive Learning for Shipboard Wind Speed Correction
Song, J. G., Huang, M., Li, X., Zhang, Z., Wang, C., & Zhao, Z. (2025) · Journal of Ocean University of China · doi:10.1007/s11802-025-5897-9 · Cited by 2
“Contrastive learning, a technique that learns feature re-presentations by comparing diverse samples (Le-Khac et al ., 2020), has gained prominence across different domains, such as computer vision, speech, and time-series analysis (Chen et al ., 2020).”
MixDiff: Mixing Natural and Synthetic Images for Robust Self-Supervised Representations
Bafghi, R. A., Harilal, N., Monteleoni, C., & Raissi, M. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00729 · Cited by 3
“Joint-embedding SSL methods can include distillation [12, 25, 76] or contrastive strategies [13, 29, 38, 81], where multiple network branches aim to learn representations by maximizing agreement between differently augmented views of the same data example in the embedding space.”
Degradation-Aware Frequency-Separated Transformer for Blind Super-Resolution
Zhao, H., Wang, B., Lu, W., & Lin, J. (2025) · Lecture notes in computer science · doi:10.1007/978-981-96-5809-1_13 · Cited by 1
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.