{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,28]],"date-time":"2026-10-28T12:06:41Z","timestamp":1793189201078,"version":"4.5.0"},"reference-count":144,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"2","license":[{"start":{"date-parts":[[2025,2,1]],"date-time":"2025-02-01T00:00:00Z","timestamp":1738368000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,2,1]],"date-time":"2025-02-01T00:00:00Z","timestamp":1738368000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,2,1]],"date-time":"2025-02-01T00:00:00Z","timestamp":1738368000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Rensselaer-IBM AI Research Collaboration"},{"name":"IBM AI Horizons Network","id":[{"id":"https:\/\/ror.org\/05mzexv92","id-type":"ROR","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Neural Netw. Learning Syst."],"published-print":{"date-parts":[[2025,2]]},"DOI":"10.1109\/tnnls.2023.3338619","type":"journal-article","created":{"date-parts":[[2023,12,13]],"date-time":"2023-12-13T14:38:30Z","timestamp":1702478310000},"page":"2133-2153","source":"Crossref","is-referenced-by-count":44,"title":["Knowledge-Augmented Deep Learning and its Applications: A Survey"],"prefix":"10.1109","volume":"36","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4362-197X","authenticated-orcid":false,"given":"Zijun","family":"Cui","sequence":"first","affiliation":[{"name":"Rensselaer Polytechnic Institute, Troy, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tian","family":"Gao","sequence":"additional","affiliation":[{"name":"IBM Research, Yorktown Heights, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4628-3785","authenticated-orcid":false,"given":"Kartik","family":"Talamadupula","sequence":"additional","affiliation":[{"name":"IBM Research, Yorktown Heights, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4302-2889","authenticated-orcid":false,"given":"Qiang","family":"Ji","sequence":"additional","affiliation":[{"name":"Department of Electrical, Computer, and Systems Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Deep learning: A critical appraisal","author":"Marcus","year":"2018","journal-title":"arXiv:1801.00631"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.3233\/faia210348"},{"key":"ref3","article-title":"A survey on neural-symbolic learning systems","author":"Yu","year":"2021","journal-title":"arXiv:2111.08164"},{"key":"ref4","article-title":"Integrating physics-based modeling with machine learning: A survey","author":"Willard","year":"2020","journal-title":"arXiv:2003.04919v4"},{"key":"ref5","article-title":"Integrating machine learning with physics-based modeling","author":"Weinan","year":"2020","journal-title":"arXiv:2006.02619"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.2987324"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-024-10722-5"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1088\/1748-9326\/aae2be"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1016\/j.isci.2020.101656"},{"key":"ref10","article-title":"Informed machine learning\u2014A taxonomy and survey of integrating knowledge into learning systems","author":"von Rueden","year":"2019","journal-title":"arXiv:1903.12394"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1007\/s12206-021-0342-5"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-44584-3_43"},{"key":"ref13","article-title":"Knowledge as invariance\u2014History and perspectives of knowledge-augmented machine learning","author":"Sagel","year":"2020","journal-title":"arXiv:2012.11406"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2013.2274748"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00995"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00443"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611976236.60"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611976236.63"},{"key":"ref19","article-title":"Achieving conservation of energy in neural network emulators for climate modeling","author":"Beucler","year":"2019","journal-title":"arXiv:1906.06622"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.126.098302"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.120.143001"},{"key":"ref22","article-title":"Controlling neural networks with rule representations","author":"Seo","year":"2021","journal-title":"arXiv:2106.07804"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1190\/geo2019-0138.1"},{"key":"ref24","article-title":"Hamiltonian generative networks","author":"Toth","year":"2019","journal-title":"arXiv:1909.13789"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.101.062207"},{"key":"ref26","article-title":"Neural networks with physics-informed architectures and constraints for dynamical systems modeling","author":"Djeumou","year":"2021","journal-title":"arXiv:2109.06407"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2018.10.045"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.1038\/s42254-021-00314-5"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00884"},{"key":"ref30","article-title":"Hamiltonian neural networks","author":"Greydanus","year":"2019","journal-title":"arXiv:1906.01563"},{"key":"ref31","article-title":"Symplectic ODE-Net: Learning Hamiltonian dynamics with control","author":"Zhong","year":"2019","journal-title":"arXiv:1909.12077"},{"key":"ref32","article-title":"Lagrangian neural networks","author":"Cranmer","year":"2020","journal-title":"arXiv:2003.04630"},{"key":"ref33","article-title":"LagNetViP: A Lagrangian neural network for video prediction","author":"Allen-Blanchette","year":"2020","journal-title":"arXiv:2010.12932"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v31i1.10934"},{"key":"ref35","article-title":"Deep imitation learning for bimanual robotic manipulation","author":"Xie","year":"2020","journal-title":"arXiv:2010.05134"},{"key":"ref36","first-page":"153","article-title":"Learning to see physics via visual de-animation","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Wu"},{"key":"ref37","first-page":"1","article-title":"Differentiable simulation of soft multi-body systems","volume-title":"Proc. 35th Conf. Neural Inf. Process. Syst.","author":"Qiao"},{"key":"ref38","article-title":"Which priors matter? Benchmarking models for learning latent dynamics","author":"Botev","year":"2021","journal-title":"arXiv:2111.05458"},{"key":"ref39","article-title":"Incorporating symmetry into deep dynamics models for improved generalization","author":"Wang","year":"2020","journal-title":"arXiv:2002.03061"},{"key":"ref40","first-page":"1","article-title":"Residual pathway priors for soft equivariance constraints","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Finzi"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1017\/jfm.2016.615"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00921"},{"key":"ref43","article-title":"Encoding invariances in deep generative models","author":"Shah","year":"2019","journal-title":"arXiv:1906.01626"},{"key":"ref44","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2019.109209"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2019.109056"},{"key":"ref46","article-title":"Cormorant: Covariant molecular neural networks","author":"Anderson","year":"2019","journal-title":"arXiv:1906.04015"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00436"},{"key":"ref48","article-title":"A compositional object-based approach to learning physical dynamics","author":"Chang","year":"2016","journal-title":"arXiv:1612.00341"},{"key":"ref49","article-title":"Interaction networks for learning about objects, relations and physics","author":"Battaglia","year":"2016","journal-title":"arXiv:1612.00222"},{"key":"ref50","article-title":"Physics-integrated variational autoencoders for robust and interpretable generative modeling","author":"Takeishi","year":"2021","journal-title":"arXiv:2102.13156"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1088\/1742-5468\/ac3ae5"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1145\/3450439.3451866"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2019.05.027"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2019.05.024"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2019.109120"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611975673.63"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1145\/3447814"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA.2018.8460875"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-67361-5_40"},{"key":"ref60","volume-title":"Pybullet, a Python Module for Physics Simulation for Games, Robotics and Machine Learning","author":"Coumans","year":"2016"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00030"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.383"},{"key":"ref63","doi-asserted-by":"publisher","DOI":"10.1109\/IROS.2017.8202133"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/T-AFFC.2013.5"},{"key":"ref65","first-page":"86","article-title":"Incorporating causal graphical prior knowledge into predictive modeling via simple data augmentation","volume-title":"Proc. Uncertainty Artif. Intell.","author":"Teshima"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01258-8_19"},{"key":"ref67","first-page":"1","article-title":"Deep structured prediction for facial landmark detection","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"32","author":"Chen"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01372"},{"key":"ref69","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00620"},{"key":"ref70","article-title":"Matching learned causal effects of neural networks with domain priors","author":"Kancheti","year":"2021","journal-title":"arXiv:2111.12490"},{"key":"ref71","first-page":"8116","article-title":"Interpretations are useful: Penalizing explanations to align neural networks with prior knowledge","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Rieger"},{"key":"ref72","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00536"},{"key":"ref73","first-page":"1","article-title":"Knowledge augmented deep neural networks for joint facial expression and action unit recognition","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"33","author":"Cui"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00947"},{"key":"ref75","article-title":"Probabilistic logic neural networks for reasoning","author":"Qu","year":"2019","journal-title":"arXiv:1906.08495"},{"key":"ref76","first-page":"5502","article-title":"A semantic loss function for deep learning with symbolic knowledge","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Xu"},{"key":"ref77","article-title":"Embedding symbolic knowledge into deep networks","author":"Xie","year":"2019","journal-title":"arXiv:1909.01161"},{"key":"ref78","first-page":"1853","article-title":"Symbolic graph reasoning meets convolutions","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"31","author":"Liang"},{"key":"ref79","first-page":"1","article-title":"Auto-encoding knowledge graph for unsupervised medical report generation","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Liu"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i07.6841"},{"key":"ref81","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2017\/230"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00207"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01389"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i8.16879"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1145\/3491209"},{"key":"ref86","first-page":"5105","article-title":"Feature importance ranking for deep learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"33","author":"Wojtas"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.1023\/b:visi.0000029664.99615.94"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1007\/s13042-010-0001-0"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1145\/3594315.3594328"},{"key":"ref90","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2013.50"},{"key":"ref91","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbaa199"},{"key":"ref92","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-30796-7_10"},{"key":"ref93","doi-asserted-by":"publisher","DOI":"10.1016\/j.patrec.2017.01.002"},{"key":"ref94","article-title":"Physics informed deep kernel learning","author":"Wang","year":"2020","journal-title":"arXiv:2006.04976"},{"key":"ref95","doi-asserted-by":"publisher","DOI":"10.1021\/acs.jctc.8b00025"},{"key":"ref96","article-title":"SE(3)-equivariant prediction of molecular wavefunctions and electronic densities","author":"Unke","year":"2021","journal-title":"arXiv:2106.02347"},{"key":"ref97","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.00840"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00708"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.1145\/3450626.3459825"},{"key":"ref100","doi-asserted-by":"publisher","DOI":"10.1126\/science.177.4047.393"},{"key":"ref101","doi-asserted-by":"publisher","DOI":"10.1017\/cbo9781139644181"},{"key":"ref102","doi-asserted-by":"publisher","DOI":"10.1007\/BF01679684"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.1145\/15922.15902"},{"key":"ref104","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00487"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.1109\/AMFG.2003.1240831"},{"key":"ref106","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00696"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00521"},{"key":"ref108","first-page":"1","article-title":"Learning to dehaze with polarization","volume-title":"Proc. 35th Conf. Neural Inf. Process. Syst.","author":"Zhou"},{"key":"ref109","first-page":"1","article-title":"Learning with algorithmic supervision via continuous relaxations","volume-title":"Proc. 35th Conf. Neural Inf. Process. Syst.","author":"Petersen"},{"key":"ref110","doi-asserted-by":"publisher","DOI":"10.1017\/cbo9780511811685"},{"key":"ref111","article-title":"Deep Lagrangian networks: Using physics as model prior for deep learning","author":"Lutter","year":"2019","journal-title":"arXiv:1907.04490"},{"key":"ref112","article-title":"Combining generative and discriminative models for hybrid inference","author":"Satorras","year":"2019","journal-title":"arXiv:1906.02547"},{"key":"ref113","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01809"},{"key":"ref114","doi-asserted-by":"publisher","DOI":"10.1109\/TRO.2020.2988642"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-46493-0_17"},{"key":"ref116","first-page":"127","article-title":"Galileo: Perceiving physical object properties by integrating a physics engine with deep learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"28","author":"Wu"},{"key":"ref117","doi-asserted-by":"publisher","DOI":"10.1109\/IROS.2012.6386109"},{"issue":"593","key":"ref118","first-page":"259","article-title":"An article on optics of paint layers","volume":"12","author":"Kubelka","year":"1931","journal-title":"Z. Tech. Phys."},{"key":"ref119","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2021.3070878"},{"key":"ref120","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00828"},{"key":"ref121","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2022.3140726"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-03778-8_11"},{"key":"ref123","doi-asserted-by":"publisher","DOI":"10.1109\/TCSVT.2021.3057267"},{"key":"ref124","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00738"},{"key":"ref125","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.121"},{"key":"ref126","first-page":"1","article-title":"Robust visual reasoning via language guided neural module networks","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Akula"},{"key":"ref127","first-page":"1","article-title":"Leveraging the inductive bias of large language models for abstract textual reasoning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Rytting"},{"key":"ref128","first-page":"1","article-title":"Refining language models with compositional explanations","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Yao"},{"key":"ref129","article-title":"Facial action coding system","author":"Ekman","year":"1978","journal-title":"Environ. Psychol. Nonverbal Behav."},{"key":"ref130","volume-title":"A Mathematical Introduction to Logic","author":"Enderton","year":"2001"},{"key":"ref131","doi-asserted-by":"publisher","DOI":"10.1109\/CSCWD.2019.8791922"},{"key":"ref132","doi-asserted-by":"publisher","DOI":"10.1162\/dint_a_00013"},{"key":"ref133","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW59228.2023.00108"},{"key":"ref134","doi-asserted-by":"publisher","DOI":"10.1016\/0004-3702(94)90105-8"},{"key":"ref135","doi-asserted-by":"publisher","DOI":"10.1023\/A:1008328630915"},{"key":"ref136","article-title":"Logical neural networks","author":"Riegel","year":"2020","journal-title":"arXiv:2006.13155"},{"key":"ref137","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2014.2327636"},{"key":"ref138","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2017\/221"},{"key":"ref139","doi-asserted-by":"publisher","DOI":"10.1016\/j.artint.2015.08.011"},{"key":"ref140","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v31i1.11164"},{"key":"ref141","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijar.2012.10.004"},{"key":"ref142","volume-title":"Constructing Bayesian Networks for Linear Dynamic Systems","author":"Evers","year":"2011"},{"key":"ref143","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2020.3017010"},{"key":"ref144","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2016.2603784"}],"container-title":["IEEE Transactions on Neural Networks and Learning Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/5962385\/10877690\/10359123.pdf?arnumber=10359123","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,12,5]],"date-time":"2025-12-05T18:39:01Z","timestamp":1764959941000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10359123\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,2]]},"references-count":144,"journal-issue":{"issue":"2"},"URL":"https:\/\/doi.org\/10.1109\/tnnls.2023.3338619","relation":{},"ISSN":["2162-237X","2162-2388"],"issn-type":[{"value":"2162-237X","type":"print"},{"value":"2162-2388","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,2]]}}}