{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,17]],"date-time":"2026-03-17T23:26:08Z","timestamp":1773789968051,"version":"3.50.1"},"reference-count":59,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"3","license":[{"start":{"date-parts":[[2020,3,1]],"date-time":"2020-03-01T00:00:00Z","timestamp":1583020800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"name":"Fraunhofer Society through the MPI-FhG collaboration project \u201cTheory and Practice for Reduced Learning Machines,\u201d"},{"DOI":"10.13039\/501100002347","name":"Bundesministerium f\u00fcr Bildung und Forschung","doi-asserted-by":"publisher","award":["01IS14013A"],"award-info":[{"award-number":["01IS14013A"]}],"id":[{"id":"10.13039\/501100002347","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002347","name":"Bundesministerium f\u00fcr Bildung und Forschung","doi-asserted-by":"publisher","award":["01IS18037I"],"award-info":[{"award-number":["01IS18037I"]}],"id":[{"id":"10.13039\/501100002347","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001659","name":"Deutsche Forschungsgemeinschaft","doi-asserted-by":"publisher","award":["390685689"],"award-info":[{"award-number":["390685689"]}],"id":[{"id":"10.13039\/501100001659","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Information and Communications Technology Planning and Evaluation (IITP) Grant"},{"name":"Korean Government","award":["2017-0-00451"],"award-info":[{"award-number":["2017-0-00451"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Neural Netw. Learning Syst."],"published-print":{"date-parts":[[2020,3]]},"DOI":"10.1109\/tnnls.2019.2910073","type":"journal-article","created":{"date-parts":[[2019,5,29]],"date-time":"2019-05-29T19:38:42Z","timestamp":1559158722000},"page":"772-785","source":"Crossref","is-referenced-by-count":59,"title":["Compact and Computationally Efficient Representation of Deep Neural Networks"],"prefix":"10.1109","volume":"31","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5144-3758","authenticated-orcid":false,"given":"Simon","family":"Wiedemann","sequence":"first","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, Berlin, Germany"}]},{"given":"Klaus-Robert","family":"Muller","sequence":"additional","affiliation":[{"name":"Technische Universit\u00e4t Berlin, Berlin, Germany"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6283-3265","authenticated-orcid":false,"given":"Wojciech","family":"Samek","sequence":"additional","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, Berlin, Germany"}]}],"member":"263","reference":[{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.2019.8852119"},{"key":"ref38","author":"federici","year":"2017","journal-title":"Improved Bayesian compression"},{"key":"ref33","author":"li","year":"2016","journal-title":"Ternary Weight Networks"},{"key":"ref32","first-page":"2849","article-title":"Fixed point quantization of deep convolutional networks","author":"lin","year":"2016","journal-title":"Proc Int Conf Mach Learn (ICML)"},{"key":"ref31","first-page":"1","article-title":"Improving the speed of neural networks on CPUs","author":"vanhoucke","year":"2011","journal-title":"Proc NIPS Workshop on Deep Learning and Unsupervised Feature Learning"},{"key":"ref30","first-page":"1135","article-title":"Learning both weights and connections for efficient neural network","author":"han","year":"2015","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref37","first-page":"3290","article-title":"Bayesian compression for deep learning","author":"louizos","year":"2017","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref36","author":"ullrich","year":"2017","journal-title":"Soft weight-sharing for neural network compression"},{"key":"ref35","author":"choi","year":"2018","journal-title":"Universal deep neural network compression"},{"key":"ref34","author":"choi","year":"2016","journal-title":"Towards the Limit of Network Quantization"},{"key":"ref28","first-page":"598","article-title":"Optimal brain damage","author":"lecun","year":"1990","journal-title":"Advances in neural information processing systems"},{"key":"ref27","author":"cheng","year":"2017","journal-title":"A survey of model compression and acceleration for deep neural networks"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ICNN.1993.298572"},{"key":"ref2","author":"young","year":"1988","journal-title":"A Survey of Numerical Mathematics"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1515\/9781400841189"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1016\/j.jneumeth.2016.10.008"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.1603015"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms13890"},{"key":"ref24","author":"hinton","year":"2015","journal-title":"Distilling the knowledge in a neural network"},{"key":"ref23","first-page":"2148","article-title":"Predicting parameters in deep learning","author":"denil","year":"2013","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref26","first-page":"2498","article-title":"Variational dropout sparsifies deep neural networks","author":"molchanov","year":"2017","journal-title":"Proc Int Conf Mach Learn (ICML)"},{"key":"ref25","author":"han","year":"2015","journal-title":"Deep compression Compressing deep neural networks with pruning trained quantization and huffman coding"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.643"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1109\/JSSC.2016.2616357"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-019-08987-4"},{"key":"ref58","first-page":"39","article-title":"Explainable artificial intelligence: Understanding, visualizing and interpreting deep learning models","volume":"1","author":"samek","year":"2018","journal-title":"ITU J ICT Discoveries"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.2019.8852172"},{"key":"ref56","author":"mcmahan","year":"2016","journal-title":"Communication-efficient learning of deep networks from decentralized data"},{"key":"ref55","first-page":"1097","article-title":"ImageNet classification with deep convolutional neural networks","author":"krizhevsky","year":"2012","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1145\/3007787.3001163"},{"key":"ref53","first-page":"4700","article-title":"Densely connected convolutional networks","author":"huang","year":"2016","journal-title":"Proc IEEE Conf Comput Vis Pattern Recognit (CVPR)"},{"key":"ref52","first-page":"10","article-title":"1.1 computing’s energy problem (and what we can do about it)","author":"horowitz","year":"2014","journal-title":"IEEE Int Solid-State Circuits Conf (ISSCC) Dig Tech Papers"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-30140-0_54"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/584091.584093"},{"key":"ref40","author":"simonyan","year":"2014","journal-title":"Very Deep Convolutional Networks for Large-scale Image Recognition"},{"key":"ref12","first-page":"547","article-title":"On measures of entropy and information","volume":"1","author":"r\u00e9nyi","year":"1961","journal-title":"Proc 4th Berkeley Symp Math Statist Probab Contrib Theory Statist"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2014.09.003"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.123"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/TIP.2017.2760518"},{"key":"ref17","author":"dai","year":"2016","journal-title":"Very deep convolutional neural networks for raw waveforms"},{"key":"ref18","first-page":"1","article-title":"Neural machine translation by jointly learning to align and translate","author":"bahdanau","year":"2015","journal-title":"Proc Int Conf Represent Learn (ICLR)"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1038\/ncomms5308"},{"key":"ref4","doi-asserted-by":"crossref","first-page":"9","DOI":"10.1007\/978-3-642-35289-8_3","article-title":"Efficient BackProp","volume":"7700","author":"lecun","year":"2012","journal-title":"Neural Networks Tricks of the Trade"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4302-5930-5"},{"key":"ref6","first-page":"1","article-title":"Fast matrix multiplication","volume":"5","author":"bl\u00e4ser","year":"2013","journal-title":"Theory of Computing Graduate Surveys"},{"key":"ref5","first-page":"151","article-title":"Survey on matrix multiplication algorithms","author":"afroz","year":"2016","journal-title":"Proc Int Conf Inf Electron Vis (ICIEV)"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/TPDS.2012.290"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/PROC.1977.10514"},{"key":"ref49","author":"deng","year":"2017","journal-title":"GXNOR-Net Training deep neural networks with ternary weights and activations without full-precision memory under a unified discretization framework"},{"key":"ref9","doi-asserted-by":"crossref","first-page":"61","DOI":"10.1007\/978-3-319-41321-1_4","article-title":"Dynamic-CSR: A format for dynamic sparse-matrix updates","volume":"9697","author":"king","year":"2016","journal-title":"Proc 31st Int Conf High Perform Comput"},{"key":"ref46","author":"courbariaux","year":"2016","journal-title":"Binarized neural networks Training deep neural networks with weights and activations constrained to +1 or ?1"},{"key":"ref45","author":"mellempudi","year":"2017","journal-title":"Mixed Low-precision Deep Learning Inference using Dynamic Fixed Point"},{"key":"ref48","author":"rastegari","year":"2016","journal-title":"Xnor-net Imagenet classification using binary convolutional neural networks"},{"key":"ref47","author":"kim","year":"2016","journal-title":"Bitwise Neural Networks"},{"key":"ref42","author":"wang","year":"2018","journal-title":"Training deep neural networks with 8-bit floating point numbers"},{"key":"ref41","first-page":"5145","article-title":"Scalable methods for 8-bit training of neural networks","author":"banner","year":"2018","journal-title":"Proc Adv Neural Inf Process Syst"},{"key":"ref44","year":"2019","journal-title":"QNNPACK open source library for optimized mobile deep learning"},{"key":"ref43","year":"2019","journal-title":"TensorFlow Lite"}],"container-title":["IEEE Transactions on Neural Networks and Learning Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/5962385\/9018332\/08725933.pdf?arnumber=8725933","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,1,27]],"date-time":"2022-01-27T19:26:11Z","timestamp":1643311571000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/8725933\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,3]]},"references-count":59,"journal-issue":{"issue":"3"},"URL":"https:\/\/doi.org\/10.1109\/tnnls.2019.2910073","relation":{},"ISSN":["2162-237X","2162-2388"],"issn-type":[{"value":"2162-237X","type":"print"},{"value":"2162-2388","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,3]]}}}