RFID based real-time recognition of ongoing gesture with adversarial learning
The DOI matches its Crossref record; 43 of the 44 references it cites were found in a registry or an index.
Abstract
At present, wireless sensing based gesture recognition is becoming a rising star due to its convenience and non-invasiveness without privacy issues, while the strict requirement of the deployment and surrounding environment is still an unavoidable issue which limits its development and generalization. Although there are some works involving the environmental variance, the changes of relative positions between devices and users are ignored. As one of the most popular wireless sensing methods, RFID is widely used in activity recognition with its stable low-level physical characters such as phase and RSS. Besides, the signals reflected from RFID tags intuitively delineate its movements. On the other hand, many interactive gesture-driven applications, such as gesture input for video games, have a paramount and unavoidable issue about the latency between completion of a gesture and its recognition. Inspired by deep learning, this paper presents a real-time ongoing gesture recognition system EUIGR, which efficiently integrates phase and RSS data streams, and extracts both environment and user invariant features. The proposed system seamlessly integrates CNNs (Convolutional Neural Networks) and LSTM (Long Short-Term Memory) to fuse RFID low-level physical characters and extract space-temporal information. Furthermore, with adversarial learning, EUIGR suppresses environment-related factors and the user-specific features, and obtains strong robustness to individual diversity and decreases the environmental dependence. We also implement the system with COTS RFID devices, and extensive experimental results show the effectiveness and accuracy of EUIGR.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
(2008) · Journal of Machine Learning Research · Cited by 33,051
Learning long-term dependencies with gradient descent is difficult
Bengio, Y., Simard, P., & Frasconi, P. (1994) · IEEE Transactions on Neural Networks · doi:10.1109/72.279181 · Cited by 8,542 · cited 1×
“However, standard RNN has the problem of long-term dependencies due to gradient descent of an error criterion[2].”
BackgroundDomain-Adversarial Training of Neural Networks
Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., Marchand, M., & Lempitsky, V. (2017) · Advances in Computer Vision and Pattern Recognition · doi:10.1007/978-3-319-58347-1_10 · Cited by 8,318 · cited 1×
“Adversarial learning are utilized to learn robust representations which are discriminative on source domain while invariant between domains[10, 32].”
BackgroundAdversarial Discriminative Domain Adaptation
Tzeng, E., Hoffman, J., Saenko, K., & Darrell, T. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.316 · Cited by 5,154 · cited 1×
“Adversarial learning are utilized to learn robust representations which are discriminative on source domain while invariant between domains[10, 32].”
BackgroundLANDMARC: indoor location sensing using active RFID
Ni, L. M., Liu, Y., Lau, Y. C., & Patil, A. P. (2004) · Proceedings of the First IEEE International Conference on Pervasive Computing and Communications, 2003. (PerCom 2003). · doi:10.1109/percom.2003.1192765 · Cited by 2,323
LANDMARC: Indoor Location Sensing Using Active RFID
Ni, L. M., Liu, Y., Lau, Y. C., & Patil, A. P. (2004) · Wireless Networks · doi:10.1023/b:wine.0000044029.06344.dd · Cited by 1,420 · cited 1×
“It has also achieved great success in gesture recognition[6, 18, 29, 44] and indoor localization[25, 34, 37] using RSS or phase information reported by COTS (Commercial Off-The-Shelf) RFID readers.”
MethodDomain-adversarial training of neural networks
GaninYaroslav, UstinovaEvgeniya, AjakanHana, GermainPascal, LarochelleHugo, LavioletteFrançois, MarchandMario, & LempitskyVictor (2016) · Journal of Machine Learning Research · doi:10.5555/2946645.2946704 · Cited by 1,201
Soli: ubiquitous gesture sensing with millimeter wave radar
Lien, J., Gillian, N., Karagozler, M. E., Amihood, P., Schwesig, C., Olson, E., Raja, H., & Poupyrev, I. (2016) · ACM Transactions on Graphics · doi:10.1145/2897824.2925953 · Cited by 957
Analysis of the phase unwrapping algorithm
Itoh, K. (1982) · Applied Optics · doi:10.1364/ao.21.002470 · Cited by 887 · cited 1×
“Accordingly, we adopt an extensively applied phase unwrapping method[17].”
MethodTagoram: real-time tracking of mobile RFID tags to high precision using COTS devices
Yang, L., Chen, Y., Li, X.-Y., Xiao, C., Li, M., & Liu, Y. (2014) · Proceedings of the 20th annual international conference on Mobile computing and networking · doi:10.1145/2639108.2639111 · Cited by 757 · cited 1×
“It has also achieved great success in gesture recognition[6, 18, 29, 44] and indoor localization[25, 34, 37] using RSS or phase information reported by COTS (Commercial Off-The-Shelf) RFID readers.”
MethodMolchanov, P., Yang, X., Gupta, S., Kim, K., Tyree, S., & Kautz, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.456 · Cited by 695 · cited 3×
“Furthermore, Section 4 illustrates the detailed design of the gesture recognition system with negative delay[14] and EUIT.”
BackgroundDude, where's my card?: RFID positioning that works with multipath and non-line of sight
Wang, J., & Katabi, D. (2013) · Proceedings of the ACM SIGCOMM 2013 conference on SIGCOMM · doi:10.1145/2486001.2486029 · Cited by 482 · cited 1×
“It has also achieved great success in gesture recognition[6, 18, 29, 44] and indoor localization[25, 34, 37] using RSS or phase information reported by COTS (Commercial Off-The-Shelf) RFID readers.”
Method2015 IEEE Conference on Computer Communications (INFOCOM)
(2015) · doi:10.1109/infocom30874.2015 · Cited by 410
SignFi: Sign Language Recognition Using WiFi
Ma, Y., Zhou, G., Wang, S., Zhao, H., & Jung, W. (2018) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3191755 · Cited by 380 · cited 3×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundDevice-free gesture tracking using acoustic signals
Wang, W., Liu, A. X., & Sun, K. (2016) · Proceedings of the 22nd Annual International Conference on Mobile Computing and Networking · doi:10.1145/2973750.2973764 · Cited by 378 · cited 1×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundAdversarial Discriminative Domain Adaptation
Tzeng, E., Hoffman, J., Saenko, K., & Darrell, T. J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1702.05464 · Cited by 363 · cited 1×
“Adversarial learning are utilized to learn robust representations which are discriminative on source domain while invariant between domains[10, 32].”
BackgroundRF-IDraw: virtual touch screen in the air using RF signals
Wang, J., Vasisht, D., & Katabi, D. (2014) · Proceedings of the 2014 ACM conference on SIGCOMM · doi:10.1145/2619239.2626330 · Cited by 359
Cui, R., Liu, H., & Zhang, C. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.175 · Cited by 338 · cited 1×
“Most recently, researchers have widely exploited Leap Motion or Kinect to build gesture recognition systems to enable sign language translation at both word and sentence levels with high accuracy[9], to recognize continuous sign language with weakly supervised learning[5].”
BackgroundRisQ: recognizing smoking gestures with inertial sensors on a wristband
Parate, A., Chiu, M.-C., Chadowitz, C., Ganesan, D., & Kalogerakis, E. (2014) · Proceedings of the 12th annual international conference on Mobile systems, applications, and services · doi:10.1145/2594368.2594379 · Cited by 282 · cited 2×
“For example, wrist-worn devices containing inertial sensors are utilized to recognize the eating gesture[28] and identify smoking gestures[27], translate sign language[39, 41], and identify fine-grained interactive gestures[12], respectively.”
BackgroundMax-Margin Early Event Detectors
Hoai, M., & De la Torre, F. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-013-0683-3 · Cited by 212 · cited 3×
“Hoai et al.[15] design computer vision based early event detectors which enable sign language translation, emotions recognition, respectively.”
BackgroundMethodFEMO: A Platform for Free-weight Exercise Monitoring with RFIDs
Ding, H., Shangguan, L., Yang, Z., Han, J., Zhou, Z., Yang, P., Xi, W., & Zhao, J. (2015) · Proceedings of the 13th ACM Conference on Embedded Networked Sensor Systems · doi:10.1145/2809695.2809708 · Cited by 183 · cited 2×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundMethodGRfid: A Device-Free RFID-Based Gesture Recognition System
Zou, Y., Xiao, J., Han, J., Wu, K., Li, Y., & Ni, L. M. (2017) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2016.2549518 · Cited by 165 · cited 5×
“It has also achieved great success in gesture recognition[6, 18, 29, 44] and indoor localization[25, 34, 37] using RSS or phase information reported by COTS (Commercial Off-The-Shelf) RFID readers.”
BackgroundMethodInfluentialSpeaker-Invariant Training Via Adversarial Learning
Meng, Z., Li, J., Chen, Z., Zhao, Y., Mazalov, V., Gong, Y., & Juang, B.-H. F. (2018) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2018.8461932 · Cited by 139 · cited 1×
“SIT [24] utilizes multi-task optimization to reduce variance across speakers and keep representations senone-discriminative.”
Recognizing Keystrokes Using WiFi Devices
Ali, K., Liu, A. X., Wang, W., & Shahzad, M. (2017) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2017.2680998 · Cited by 117 · cited 2×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundWristWhirl: One-handed Continuous Smartwatch Input using Wrist Gestures
Gong, J., Yang, X.-D., & Irani, P. (2016) · Proceedings of the 29th Annual Symposium on User Interface Software and Technology · doi:10.1145/2984511.2984563 · Cited by 112 · cited 2×
“For example, wrist-worn devices containing inertial sensors are utilized to recognize the eating gesture[28] and identify smoking gestures[27], translate sign language[39, 41], and identify fine-grained interactive gestures[12], respectively.”
BackgroundDeepASL: Enabling Ubiquitous and Non-Intrusive Word and Sentence-Level Sign Language Translation
Fang, B., Co, J., & Zhang, M. (2017) · Proceedings of the 15th ACM Conference on Embedded Network Sensor Systems · doi:10.1145/3131672.3131693 · Cited by 107 · cited 2×
“In addition, a number of recent works on gesture detection are designed for recognizing gesture after it is completed[9, 22], so the latency between gestures completion and recognition cannot be avoided.”
BackgroundEnabling Gesture-based Interactions with Objects
Shangguan, L., Zhou, Z., & Jamieson, K. (2017) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3081333.3081364 · Cited by 101 · cited 3×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundMethodShangguan, L., Zhou, Z., Zheng, X., Yang, L., Liu, Y., & Han, J. (2015) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/2809695.2809710 · Cited by 95 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundChinese Sign Language Recognition Based on an Optimized Tree-Structure Framework
Yang, X., Chen, X., Cao, X., Wei, S., & Zhang, X. (2017) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2016.2560907 · Cited by 93 · cited 3×
“…human and computer compared with the text-based and graphic user interface, which draws great attention in recent years with the rapid development of the artificial intelligence, mobile computing, Internet of Things, virtual reality, argument reality and other technologies[1, 22, 29, 35, 39].”
BackgroundUltraGesture: Fine-Grained Gesture Sensing and Recognition
Ling, K., Dai, H., Liu, Y., & Liu, A. X. (2018) · IEEE International Conference on Sensing, Communication, and Networking (SECON) · doi:10.1109/sahcn.2018.8397099 · Cited by 83 · cited 1×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundZhang, Q., Wang, D., Zhao, R., & Yu, Y. (2019) · ACM International Conference on Intelligent User Interfaces (IUI) · doi:10.1145/3301275.3302296 · Cited by 78 · cited 1×
“For example, wrist-worn devices containing inertial sensors are utilized to recognize the eating gesture[28] and identify smoking gestures[27], translate sign language[39, 41], and identify fine-grained interactive gestures[12], respectively.”
Self-citationTagFree Activity Identification with RFIDs
Fan, X., Gong, W., & Liu, J. (2018) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3191739 · Cited by 62 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundImproving the Recognition of Eating Gestures Using Intergesture Sequential Dependencies
Ramos-Garcia, R. I., Muth, E. R., Gowdy, J. N., & Hoover, A. W. (2014) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2014.2329137 · Cited by 57 · cited 2×
“For example, wrist-worn devices containing inertial sensors are utilized to recognize the eating gesture[28] and identify smoking gestures[27], translate sign language[39, 41], and identify fine-grained interactive gestures[12], respectively.”
BackgroundCRH: A Contactless Respiration and Heartbeat Monitoring System with COTS RFID Tags
Zhao, R., Wang, D., Zhang, Q., Chen, H., & Huang, A. (2018) · IEEE International Conference on Sensing, Communication, and Networking (SECON) · doi:10.1109/sahcn.2018.8397132 · Cited by 54 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundSelf-citationRobust Speech Recognition Using Generative Adversarial Networks
Sriram, A., Jun, H., Gaur, Y., & Satheesh, S. (2018) · 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2018.8462456 · Cited by 52
TagBooth: Deep shopping data acquisition powered by RFID tags
Liu, T., Yang, L., Li, X.-Y., Huang, H., & Liu, Y. (2015) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom.2015.7218547 · Cited by 42 · cited 1×
“Fortunately, the tag reading rate is high enough and the speed of a human gesture is limited, so the difference between two adjacent phase values should be smaller than π [21].”
BackgroundVu, T. H., Misra, A., Roy, Q., Choo, K. T. W., & Lee, Y. (2018) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3191771 · Cited by 41
RFID Light Bulb: Enabling Ubiquitous Deployment of Interactive RFID Systems
Gummeson, J., McCann, J., Yang, C. J., Ranasinghe, D. C., Hudson, S. E., & Sample, A. P. (2017) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3090077 · Cited by 40 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundViType: A Cost Efficient On-Body Typing System through Vibration
Chen, W., Guan, M., Huang, Y., Wang, L., Ruby, R., Hu, W. S., & Wu, K. (2018) · IEEE International Conference on Sensing, Communication, and Networking (SECON) · doi:10.1109/sahcn.2018.8397098 · Cited by 38 · cited 1×
“Then unlike recent works [3, 4, 44] which combine segmentation and classification stages, extract the whole gesture and then classify it, we develop an LSTM based sequence labeling classifier, where the probability of each gesture is given at every time clip.”
BackgroundZhang, Q., Li, D., Zhao, R., Wang, D., Deng, Y., & Chen, B. (2018) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom.2018.8444599 · Cited by 21 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundSelf-citationPerceiving the Slightest Tag Motion beyond Localization
Yang, L., Guo, Y., Liu, T., Wang, C., & Liu, Y. (2015) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2015.2393859 · Cited by 16 · cited 1×
“Passive RFID (Radio Frequency Identification) is widely employed in an ocean of real-world applications[8, 30, 38, 40, 43] due to low-cost, mini-size and battery-free characters[13].”
BackgroundSGRS: A sequential gesture recognition system using COTS RFID
Chen, B., Zhang, Q., Zhao, R., Li, D., & Wang, D. (2018) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc.2018.8376998 · Cited by 14 · cited 2×
“Wireless based: In recent years, wireless signals such as ultrasound, WiFi, RFID, mmWave are widely used for fine-grained inter-activegesturesrecognition[19,20,36],signlanguagetranslation[22], keystrokesrecognition[1],andlimb-levelgesturesdetection[3,6, 29,44].”
BackgroundSelf-citationRFlow-ID: Unobtrusive Workflow Recognition with COTS RFID
Zhang, J., Zhang, Q., Li, D., Zhao, R., & Wang, D. (2017) · EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services (MobiQuitous) · doi:10.1145/3144457.3144463 · Cited by 3 · cited 2×
“Although exhibiting great potentials, RFID based gesture recognition methods can handle gestures due to dynamic environments and degrade significantly when it comes to user diversity and environment variety[18, 44].”
BackgroundMethodSelf-citation
Research works printed but not linked: 4
- Global EPC, 2008, EPC Radio-Frequency Identity Protocols Class-1 Generation-2 UHF RFID Protocol for Communications at 860 MHz-960 MHz, Version, 1, 23
- Jinshi Zhang, 2017, 14th EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services. EAI, 1--10
- Wang Wei, 2016, Proceedings of the 22nd Annual International Conference on Mobile Computing and Networking. ACM, 82--94
- Zhao Mingmin, 2017, International Conference on Machine Learning. 4100--4109
Cited by
A review and categorization of techniques on device-free human activity recognition
Hussain, Z., Sheng, Q. Z., & Zhang, W. E. (2020) · Journal of Network and Computer Applications · doi:10.1016/j.jnca.2020.102738 · Cited by 241
Xu, H., Zhou, P., Tan, R., Li, M., & Shen, G. (2021) · Proceedings of the 19th ACM Conference on Embedded Networked Sensor Systems · doi:10.1145/3485730.3485937 · Cited by 150
“However, most models [13, 23, 41, 51, 52] are trained with a large amount of labeled samples and require great manual labeling efforts, which may not be practical in reality.”
BackgroundOneFi: One-Shot Recognition for Unseen Gesture via COTS WiFi
Xiao, R., Liu, J., Han, J., & Ren, K. (2021) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3485730.3485936 · Cited by 141
“While existing works usually consider only 6-8 gesture classes [39, 40, 52], we want to explore the impact of the increasing number of unseen gesture classes on our system.”
BackgroundDeep Learning for Radio-Based Human Sensing: Recent Advances and Future Directions
Nirmal, I., Khamis, A., Hassan, M., Hu, W., & Zhu, X. (2021) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2021.3058333 · Cited by 110
Artificial Intelligence of Things: A Survey
Siam, M. S. I., Ahn, H., Liu, L., Alam, S., Shen, H., Cao, Z., Shroff, N. B., Krishnamachari, B., Srivastava, M. B., & Zhang, M. (2025) · ACM Transactions on Sensor Networks · doi:10.1145/3690639 · Cited by 123
Metadata
- Authors
- Yu, Y., Wang, D., Zhao, R., & Zhang, Q.
- Affiliations
- Shanghai Jiao Tong University (CN)
- Venue
- Proceedings of the 17th Conference on Embedded Networked Sensor Systems
- Publisher
- ACM
- Published
- 2019-11-10
- Pages
- 298-310
- OpenAlex
- W2984081625
- Cited by
- 68
- References
- 44
- Citations by year
- 2026: 2 · 2025: 9 · 2024: 14 · 2023: 10 · 2022: 17 · 2021: 12 · 2020: 4
- FWCI
- 6.68
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM Conference on Embedded Networked Sensor Systems (SenSys) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Human-Computer Interaction › Hand Gesture Recognition Systems
- Physical Sciences › Computer Science › Computer Networks and Communications › RFID technology advancements
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Keywords
- RFID, wireless sensing, real-time recognition, adversarial learning, CNN-LSTM hybrid model, environmental robustness, activity recognition
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.