Surface Defect Detection Methods for Industrial Products: A Review
Open accessThe DOI matches its Crossref record; 130 of the 131 references it cites were found in a registry or an index.
Abstract
The comprehensive intelligent development of the manufacturing industry puts forward new requirements for the quality inspection of industrial products. This paper summarizes the current research status of machine learning methods in surface defect detection, a key part in the quality inspection of industrial products. First, according to the use of surface features, the application of traditional machine vision surface defect detection methods in industrial product surface defect detection is summarized from three aspects: texture features, color features, and shape features. Secondly, the research status of industrial product surface defect detection based on deep learning technology in recent years is discussed from three aspects: supervised method, unsupervised method, and weak supervised method. Then, the common key problems and their solutions in industrial surface defect detection are systematically summarized; the key problems include real-time problem, small sample problem, small target problem, unbalanced sample problem. Lastly, the commonly used datasets of industrial surface defects in recent years are more comprehensively summarized, and the latest research methods on the MVTec AD dataset are compared, so as to provide some reference for the further research and development of industrial surface defect detection technology.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Texture feature-based methods are the most common methods in the field of steel defect detection, which reflects the homogeneity phenomenon in the image and can reflect the organization and arrangement characteristics of the image surface through the grayscale distribution of pixels and their nearby spatial neighborhood [44].”
From Steel Surface Defect Recognition: A Survey (2022) · cites this paper as [44]
Citations
Cites
Cascade R-CNN: Delving Into High Quality Object Detection
Cai, Z., & Vasconcelos, N. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00644 · Cited by 7,370
Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery
Schlegl, T., Seeböck, P., Waldstein, S. M., Schmidt-Erfurth, U., & Langs, G. (2017) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-59050-9_12 · Cited by 2,627
MVTec AD — A Comprehensive Real-World Dataset for Unsupervised Anomaly Detection
Bergmann, P., Fauser, M., Sattlegger, D., & Steger, C. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00982 · Cited by 2,203 · cited 2×
“Samples taken from the MVTec AD dataset [13]: the first row: leather; the”
MethodStacked Convolutional Auto-Encoders for Hierarchical Feature Extraction
Masci, J., Meier, U., Cireşan, D., & Schmidhuber, J. (2011) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-21735-7_7 · Cited by 1,955 · cited 1×
“Literature [63] proposed a convolutional autoencoder (CAE) for unsuper‐ vised feature learning.”
BackgroundTowards Total Recall in Industrial Anomaly Detection
Roth, K., Pemula, L., Zepeda, J., Scholkopf, B., Brox, T., & Gehler, P. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01392 · Cited by 1,810
Automatic Road Crack Detection Using Random Structured Forests
Shi, Y., Cui, L., Qi, Z., Meng, F., & Chen, Z. (2016) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2016.2552248 · Cited by 1,552 · cited 1×
“com/Iskysir/Bridge_Crack_Image_Data (accessed on 21 June 2021) Crack Forest Datasets [132,133] Road crack https://github.”
Backgroundf-AnoGAN: Fast unsupervised anomaly detection with generative adversarial networks
Schlegl, T., Seeböck, P., Waldstein, S. M., Langs, G., & Schmidt-Erfurth, U. (2019) · Medical Image Analysis · doi:10.1016/j.media.2019.01.010 · Cited by 1,545 · cited 1×
“In 2019, literature [93] improved AnoGAN and proposed the f-AnoGAN model.”
BackgroundRoad crack detection using deep convolutional neural network
Zhang, L., Yang, F., Daniel Zhang, Y., & Zhu, Y. J. (2016) · 2016 IEEE International Conference on Image Processing (ICIP) · doi:10.1109/icip.2016.7533052 · Cited by 1,496
PaDiM: A Patch Distribution Modeling Framework for Anomaly Detection and Localization
Defard, T., Setkov, A., Loesch, A., & Audigier, R. (2021) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-68799-1_35 · Cited by 1,415
An End-to-End Steel Surface Defect Detection Approach via Fusing Multiple Hierarchical Features
He, Y., Song, K., Meng, Q., & Yan, Y. (2020) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2019.2915404 · Cited by 1,341
Research works printed but not linked: 18
- Ma, 2017, Adaptive segmentation algorithm for metal surface defects, Chin. J. Sci. Instrum., 38, 245
- Li, 2020, Review of Development and Application of Defect Detection Technology, Acta Autom. Sin., 46, 2319
- Zhang, 2020, Deep Learning for Real-time Applications: A Survey, J. Softw., 31, 2654
- Yu, 2020, Visual inspection of surface defects based on lightweight reconstruction network, Acta Autom. Sin., 41, 1
- Liu, 2014, Research and development on deep learning, Appl. Res. Comput., 31, 1921
- Tao, 2020, A Survey of Surface Defect Detection Methods Based on Deep Learning, Acta Autom. Sin., 47, 1017
- Tang, 2017, Review of surface defect detection based on machine vision, J. Image Graph., 22, 1640
- Feng, 2015, Mixed-state MRF detection algorithm for ink-jet printing motion texture, J. Zhejiang Univ. (Eng. Sci.), 49, 1642
- Gao, 2017, Magneto-optical imaging fractal feature detection for welding defects of high-strength steel, Trans. China Weld. Inst., 38, 1
- Li, 2020, Research on Defect Detection Algorithm of Ceramic Tile Surface with Multi-feature Fusion, Comput. Eng. Appl., 56, 191
- Kim, M.S., Park, T., and Park, P. (2019, January 9–12). Classification of Steel Surface Defect Using Convolutional Neural Network with Few Images. Proceedings of the 12th Asian Control Conference (ASCC), Kitakyusyu International Conference Center, Fukuoka, Japan.
- Cohen, N., and Hoshen, Y. (2020). Sub-Image Anomaly Detection with Deep Pyramid Correspondences. arXiv.
- Deitsch, 2019, Automatic classification of defective photovoltaic module cells in electroluminescence images, Sol. Energy, 185, 455, 10.1016/j.solener.2019.02.067 · DOI · Paper page
- Tang, S., He, F., Huang, X., and Yang, J. (2019). Online PCB Defect Detector on A New PCB Defect Dataset. arXiv.
- Li, 2019, Research on Detection Algorithm for Bridge Cracks Based on Deep Learning, Acta Autom. Sin., 45, 1727
- Yang, J., Shi, Y., and Qi, Z. (2020). DFR: Deep Feature Reconstruction for Unsupervised Anomaly Segmentation. arXiv.
- Liu, Y., Zhuang, C., and Lu, F. (2021). Unsupervised Two-Stage Anomaly Detection. arXiv.
- Wang, G., Han, S., Ding, E., and Huang, D. (2021). Student-Teacher Feature Pyramid Matching for Unsupervised Anomaly Detection. arXiv.
Cited by
Items 401–456 of 456
The Analysis of Creative Product Design Under Deep Style Transfer Algorithm
Chen, S. (2026) · International Journal of Information Technologies and Systems Approach · doi:10.4018/ijitsa.409330
Scan Path Optimization and YOLO-Based Detection for Defect Inspection of Curved and Glossy Surfaces
Kim, M.-G., Okwuosa, C. N., & Hur, J.-W. (2026) · Sensors · doi:10.3390/s26103026
“In these environments, surface curvature and reflectivity often generate irregular lighting reflections, glare, and low image contrast, which significantly reduce the reliability of conventional vision-based inspection methods [3].”
Zhang, Z., Zou, Y., Liu, X., & Zhang, X. (2026) · Scientific Reports · doi:10.1038/s41598-026-43386-y
Li, H., Wang, Y., Su, D., Jing, J., & Wu, T. (2026) · Communications in computer and information science · doi:10.1007/978-981-95-8978-4_15
Gypsum Board Defect Detection Based on Lightweight Networks
Zhao, F., Wang, Y., Jing, J., & Su, D. (2026) · Communications in computer and information science · doi:10.1007/978-981-95-8978-4_14
Self-citationZhao, N., & Chen, S. (2026) · Sensors · doi:10.3390/s26113321 · Cited by 2
Lavanya, M., Gunasundari, C., Wong, W. K., & Ng, P. K. (2026) · International Journal of Computational Intelligence Systems · doi:10.1007/s44196-026-01397-7 · Cited by 1
Mou, Z., Dai, Y., Cao, Z., Lv, Z., Wang, L. N., Shen, Y., & Fu, G. (2026) · Sensors · doi:10.3390/s26123662
“Steel, as a core fundamental material in industrial production, is widely applied in key fields such as infrastructure construction, automotive manufacturing and aerospace, and its surface quality directly determines the reliability, safety and service life of products [1–3].”
An Ultra-High-Resolution Detection Framework for Micro-Scale PCB Surface Defects
Al-E’mari, S., Sanjalawe, Y. K., & Fataftah, F. (2026) · International Conference on Computational Intelligence Approaches and Applications (ICCIAA) · doi:10.1109/icciaa68481.2026.11543540
“Most architectural designs still rely on multi-scale feature pyramids, which are intended to detect objects at varying scales, thereby diluting the vital granular structure necessary for the localization of micro-scale defects [7], [8].”
Bionic-inspired intelligent detection of surface defects on semiconductor silicon wafers
Cheng, W., Li, Z., Yu, J., Sun, F., Lyu, G., & Zhang, X. (2026) · Measurement Science and Technology · doi:10.1088/1361-6501/ae7aa3
“…deep neural network architectures—including convolutional neural networks (CNNs), generative adversarial networks (GANs), and autoencoders—have been widely adopted in industrial AOI systems due to their strong capability in handling multi-class, multi-label, and nonlinear defect features [10–12].”
Huang, D., guo, G., Yang, J., Wang, D., & 廖琳琳 (2026) · Research Square · doi:10.21203/rs.3.rs-8234620/v1
Sabuncu, M., & Avcı, S. O. (2026) · Applied Sciences · doi:10.3390/app16126010
Fuzzy Logic-Based Hybrid Deep Learning System for Metal Surface Defect Detection
Şahin, E. B., & Ari, S. (2026) · Sakarya University Journal of Computer and Information Sciences · doi:10.35377/saucis...1945069
CNS-YOLOv8: An Improved YOLOv8-Based Defect Detection Method
Geng, R., Jiang, Y., Li, J., Wu, K., Yang, Y., Li, Z. Y., & Chang, Y. (2026) · Electronics · doi:10.3390/electronics15122730
Uncertainty-Aware Industrial Defect Detection: A Unified Evaluation and Learning Framework
Liu, K., Wang, Y., & LI, Y. (2026) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc69976.2026.11559828
“With the advancement of intelligent manufacturing, vision-based automated inspection has become a core component of industrial quality control systems [1]-[2].”
Intelligent Method for Detecting Equipment Surface Damage Based on Multi-Feature Fusion
Cui, L., Han, X., Wang, F., Li, X., Zhang, F., & He, S. (2026) · Machines · doi:10.3390/machines14070728
Cascade network with detection and inpainting for low-quality phase images
Rongxin, L., Weixian, L., Sijin, 吴. W., & Renhao, L. (2026) · Optics Communications · doi:10.1016/j.optcom.2026.133541
A multi-scale cross-domain visual anomaly detection framework for real-world traffic equipment
Yu, S., Peng, J., Wu, D., Ding, F., Ren, Y., & Ma, C. (2026) · Measurement · doi:10.1016/j.measurement.2026.122503
Faster R-CNN-based surface defect detection for sheet metal inspection in intelligent manufacturing
Chen, C.-H., Maurya, S. N., Li, K.-Y., & Chang, H.-C. (2026) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-026-18383-0
Yu, J., Zhou, K., Wang, T., Gan, H., Wang, Y., & Gao, S. (2026) · Sensors · doi:10.3390/s26144597
“…in identifying and segmenting low-contrast, small-scale defects such as tiny scratches and fine cracks; third, the model adaptation cost is high—for new industrial scenarios, it is often necessary to redesign the network structure or adjust a large number of training parameters [27,28].”
A Corner-Analysis-Based Method for Dimensional Inspection and Defect Detection of Glass Chamfers
Jiang, J., Wang, T., Zhou, Y., Wang, W. H., Wu, C., & Wei, B. (2026) · Sensors · doi:10.3390/s26154764
Anomaly Detection in Manufacturing Using Classical and Quantum AI
-, M. S. N., Gupta, S., Saini, T., & Kumar, A. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-31119-1_39
Hernández-Hernández, M., & Gallegos-Melgar, A. (2026) · Metals · doi:10.3390/met16080841
Kim, M., & Yoon, H. (2026) · AI for Engineering · doi:10.3390/aieng1020008
“For example, surface defect detection has been systematically analyzed with respect to classical, shallow, and deep learning methods [10–12].”
A domain-adaptive contrastive and context-enhanced framework for steel surface defect detection
Gao, K., Wang, S., & Li, B. (2026) · Measurement Science and Technology · doi:10.1088/1361-6501/ae9570
Aramkul, S., & Sugunnasil, P. (2026) · IEEE Access · doi:10.1109/access.2026.3721567 · Cited by 1
“Defectdetection techniques are generally classified into three categories: 1) Manual inspection, based on human visual inspection, is error-prone, expensive, labor-intensive, and time-consuming [12], [13].”
Van, C. V., Tuan, N. M., Gui, A., Sun, H., Jeremy, N. H., & Jingga, K. (2026) · International Conference on Applying New Technology in Green Buildings (ATiGB) · doi:10.1109/atigb70203.2026.11628373
“YOLO-family detectors have become a standard choice for real-time industrial vision tasks owing to their single-pass inference architecture and competitive detection performance [4]–[7].”
Zhang, J., Liu, J., & Zhao, H. (2026) · Array · doi:10.1016/j.array.2026.101108
GSLA-YOLOV11: lightweight steel surface defect detector with gated multi-branch aggregation
Chen, L., & wang, Y. (2026) · Measurement Science and Technology · doi:10.1088/1361-6501/ae9976
Chen, J. (2026) · International Conference on Computer Vision, Image and Deep Learning (CVIDL) · doi:10.1109/cvidl70130.2026.11637812
“Traditional machine vision methods can improve repeatability, yet they still rely heavily on optical conditions, image quality, handcrafted features, and manually designed processing steps [6, 7].”
Chen, Y., Yu, J., Wang, T., Gan, H., Zhou, K., Wang, R., Wang, H., Gao, S., & Liu, J. (2026) · International Conference on Computer Vision, Image and Deep Learning (CVIDL) · doi:10.1109/cvidl70130.2026.11637644
“With the development of deep learning, CNN-based segmentation methods have become the mainstream approach [12].”
Self-citationA state-of-the-art review of automotive components defect detection technologies
Pang, Z., Zhang, P., Gong, L., Ma, S., Li, S., & Sun, Z. (2026) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-026-18953-2
FeLiteNet: Accurate and Compact Steel Surface Defect Segmentation for Industrial IoT Inspection
Ju, M., Ling, R., & Jin, S. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3725320
“These factors make compact steel defect segmentation substantially more di ffi cult than general semantic segmentation under cleaner visual conditions [12].”
HR-Pro: Hierarchical Residual Prototype Memory for Efficient Industrial Anomaly Detection
Feng, S., Hu, Y., Ma, Y., Zhao, Y., & He, X. (2026) · IEEE Access · doi:10.1109/access.2026.3725847
“Industrial anomaly detection has attracted increasing attention in intelligent manufacturing systems due to its important role in automated quality inspection, defect localization, and reliable production monitoring [1].”
Surface Defect Detection via Dynamic Contrastive Graph Neural Networks
Iqbal, M. A., Genlot, C., Pao, H.-K., & Supriyono, J. (2026) · IEEE International Conference on Big Data and Artificial Intelligence (BDAI) · doi:10.1109/bdai70753.2026.11655113
“Conventional defect detection methods relied on handcrafted features derived from statistical analysis, spectral filtering, and model-based segmentation, offering computational efficiency and interpretability for well-defined defects under controlled imaging conditions [3].”
A Lightweight Detail-to-Semantic Channel Network for Boiler Water-Wall Defect Classification
Zhang, S., Lu, Y., Liu, X., Geng, Z., Cheng, X., Jiang, G., Lin, X., & Han, Y. (2026) · International Conference on Unmanned Systems and Technology (UST) · doi:10.1109/ust70584.2026.11663665
“Against this background, vision-based defect inspection has emerged as a promising alternative, owing to its non-contact nature, high inspection efficiency and compatibility with automated monitoring systems [4].”
Yang, W., Yue, Y., Qian, F., Liu, C., Chu, Y., & Lin, M. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-92-2321-3_6
Dharma, F. P., Singgih, M. L., & Prastyo, D. D. (2026) · Machine Learning with Applications · doi:10.1016/j.mlwa.2026.100989
Fang, F., Sun, D., Geng, M., Qu, Z., & Gu, S. (2026) · Sensors · doi:10.3390/s26175504
“…followed by separate classification and regression heads, and one-stage methods such as the YOLO (You Only Look Once) family, which directly predict bounding boxes and class probabilities in a single forward pass, offering a favorable balance between detection accuracy and inference speed [27].”
InfluentialXiang, Z., Kang, X., Wu, W., & Zhu, T. (2026) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-026-11631-1
Development of a Lightweight Deep Learning-Based Surface Defect Detection Model
Xing, S., & Li, C. (2026) · International Conference on Machine Learning, Natural Language Processing and Modeling · doi:10.1145/3832928.3833000
“This is combined with gradient normalization and dynamic learning rate adjustment strategies for end-to-end training [6] .”
Gao, G., Li, S., Li, S., Cheng, D., Qiu, J., Han, L., & Chen, Y. (2026) · International Conference on Artificial Intelligence, Internet of Things and Cloud Computing Technology (AIoTC) · doi:10.1109/aiotc71250.2026.11688951
Self-citationCAFM: A Cross-Modal Local Alignment Fusion Method for RGB-3D Industrial Anomaly DetectionPreprint
Zhao, Y., Li, X., Zhou, S., & Zhao, Y. (2026) · Research Square · doi:10.21203/rs.3.rs-10461475/v1
MemCLR-GANomaly: An Unsupervised Anomaly Detection Approach for Resin Cut-Off Wheel Surfaces
Song, X., Zhang, R., Lv, K., Wang, L., Liu, J., & Zhu, Y. (2026) · Sensors · doi:10.3390/s26196148
“Existing wheel vision studies adopt handcrafted texture descriptors or supervised deep learning detectors for defect and wear identification [6,18].”
Arumugam, S., N., S. M., Sindhu, A., Esakkiammal, A., & Kannan, S. (2026) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.20386
Sharma, N., Nancy, V., Angappan, S. K., Ganesan, S., Tripathy, A. K., Gupta, S., Vashisth, A., Choudari, S., & Saleela, D. (2026) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.20357
“Such an evaluation can hide class-wise weaknesses and may overestimate practical suitability, especially when datasets are relatively small or visually homogeneous [12].”
Wang, D., Xie, S., Li, Y., Zhu, L., Lin, Y., & Yuan, C. (2026) · Technologies · doi:10.3390/technologies14100644
An Advanced Fabric Defect Detection Technique Using YOLOV8n
Kaushik, P., Sagnika, S., & Bilgaiyan, S. (2026) · International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC) · doi:10.1109/assic70037.2026.11717091
“These approaches can improve detection accuracy, they often increase computational complexity and resource requirements, which limits their practicality in real industrial applications [5].”
Monnet, J., Tonn, L., Ayanyan, E., Petrović, O., Herfs, W., & Brecher, C. (2026) · IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) · doi:10.1109/etfa68377.2026.11725144
Zhou, S.-H., Li, W.-T., Zhang, Y., Ye, J.-B., Ji, K.-B., Zhou, X., & Ren, Z.-H. (2027) · Applied Soft Computing · doi:10.1016/j.asoc.2026.116657
Li, W., Liu, J., Qi, X., & Zhang, W. (2026) · Aircraft Engineering and Aerospace Technology · doi:10.1108/aeat-02-2026-0053
Text-guided cascaded diffusion for feature-aligned data augmentation in PWI-based ultrasonic testing
Zhang, X., Gu, Y., Guo, Y., Niu, X., Feng, K., Song, X., Pyle, R. J., & Zhang, Y. (2026) · NDT & E International · doi:10.1016/j.ndteint.2026.103922
Yun, J., & Won, H.-I. (2026) · Journal of the korean society of manufacturing technology engineers · doi:10.7735/ksmte.2026.35.5.404
Chen, J., & Li, S. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7662004
Amarasinghe, A. G. C. S. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7669285
Huang, B., Du, G., & Yuan, H. (2026) · Sensors · doi:10.3390/s26216850
Metadata
- Authors
- Chen, Y., Ding, Y., Zhao, F., Zhang, E., Wu, Z., & Shao, L.
- Affiliations
- Xi'an University of Technology (CN)
- Corresponding author
- Chen, Yajun
- Venue
- Applied Sciences
- Publisher
- MDPI AG
- Published
- 2021-08-20
- Volume
- 11
- Issue
- 16
- Pages
- 7657
- ISSN
- 2076-3417 (ISSN-L)
- OpenAlex
- W3195969653
- Cited by
- 457
- References
- 131
- Citations by year
- 2026: 112 · 2025: 144 · 2024: 103 · 2023: 72 · 2022: 26
- FWCI
- 25.93
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Applied Sciences (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Key R&D Program of China (2018YFD0700400)
- the Scientific Research Program funded by Shaanxi Provincial Education Department (20JY053)
- the Key Research and Development Program of Shaanxi (2019 GY-080)
- Topics
- Physical Sciences › Engineering › Industrial and Manufacturing Engineering › Industrial Vision Systems and Defect Detection
- Physical Sciences › Engineering › Mechanical Engineering › Non-Destructive Testing Techniques
- Physical Sciences › Engineering › Civil and Structural Engineering › Infrastructure Maintenance and Monitoring
- Keywords
- surface defect detection, machine vision, texture features, color features, shape features, deep learning, weakly supervised learning, small sample size problem, small object detection, class imbalance, real-time detection, defect dataset
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.