{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,3]],"date-time":"2026-04-03T15:24:35Z","timestamp":1775229875204,"version":"3.50.1"},"reference-count":36,"publisher":"IEEE","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2017,3]]},"DOI":"10.1109\/icassp.2017.7953177","type":"proceedings-article","created":{"date-parts":[[2017,6,20]],"date-time":"2017-06-20T17:35:36Z","timestamp":1497980136000},"page":"5345-5349","source":"Crossref","is-referenced-by-count":24,"title":["Returnn: The RWTH extensible training framework for universal recurrent neural networks"],"prefix":"10.1109","author":[{"given":"Patrick","family":"Doetsch","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Albert","family":"Zeyer","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Paul","family":"Voigtlaender","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ilia","family":"Kulikov","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ralf","family":"Schluter","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hermann","family":"Ney","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref33","article-title":"ADADELTA: an adaptive learning rate method","volume":"abs 1212 5701","author":"zeiler","year":"2012","journal-title":"CoRR"},{"key":"ref32","first-page":"2121","article-title":"Adaptive subgradient methods for online learning and stochastic optimization","volume":"12","author":"duchi","year":"2011","journal-title":"J Mach Learn Res"},{"key":"ref31","author":"leifert","year":"2014","journal-title":"Cells in multidimensional recurrent neural networks"},{"key":"ref30","author":"graves","year":"0","journal-title":"RNNLIB A Recurrent Neural Network Library for Sequence Learning Problems"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2015.7404837"},{"key":"ref35","first-page":"1139","article-title":"On the importance of initialization and momentum in deep learning","author":"ilya","year":"2013","journal-title":"Proceedings of The 30th InternationalConference on Machine Learning"},{"key":"ref34","article-title":"Adam: A method for stochastic optimization","volume":"abs 1412 6980","author":"kingma","year":"2014","journal-title":"CoRR"},{"key":"ref10","article-title":"RASR-the RWTH Aachen university open source speech recognition toolkit","author":"rybach","year":"2011","journal-title":"IEEE ASRU"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6854207"},{"key":"ref12","author":"chollet","year":"2015","journal-title":"Keras"},{"key":"ref13","year":"0","journal-title":"Lasagne Lightweight library to build and train neural networks in Theano"},{"key":"ref14","article-title":"Blocks and Fuel: Frameworks for deep learning","author":"van merri\u00ebnboer","year":"2015","journal-title":"ArXiv Preprint"},{"key":"ref15","article-title":"TensorFlow: Large-scale machine learning on heterogeneous systems","author":"abadi","year":"2015","journal-title":"software available from tensorflow org"},{"key":"ref16","article-title":"Torch: a modular machine learning software library","author":"collobert","year":"2002","journal-title":"Tech Rep IDIAP"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/2647868.2654889"},{"key":"ref18","year":"0","journal-title":"Fast scalable easy-to-use python based deep learning framework by nervana"},{"key":"ref19","year":"2015","journal-title":"Brainstorming"},{"key":"ref28","first-page":"1223","article-title":"Large scale distributed deep networks","author":"dean","year":"2012","journal-title":"Advances in Neural Information Processing Systems 25"},{"key":"ref4","author":"bahdanau","year":"2014","journal-title":"Neural machine translation by jointly learning to align and translate"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/ICFHR.2014.54"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2017.7952599"},{"key":"ref6","article-title":"Automatic differentiation in machine learning: a survey","volume":"abs 1502 5767","author":"gunes baydin","year":"2015","journal-title":"CoRR"},{"key":"ref29","first-page":"545","article-title":"Offline handwriting recognition with multidimensional recurrent neural networks","author":"graves","year":"2008","journal-title":"Neural Information Processing Systems Foundation"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298935"},{"key":"ref8","doi-asserted-by":"crossref","DOI":"10.25080\/Majora-92bf1922-003","article-title":"Theano: a CPU and GPU math expression compiler","author":"bergstra","year":"2010","journal-title":"Proceedings of the Python for Scientific Computing Conference (SciPy)"},{"key":"ref7","article-title":"Theano: new features and speed improvements","author":"bastien","year":"2012","journal-title":"NIPS 2012 Workshop on Deep Learning and Unsupervised Feature Learning"},{"key":"ref2","author":"sak","year":"2014","journal-title":"Long short-term memory based recurrent neural network architectures for large vocabulary speech recognition"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/2647868.2654889"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"ref20","article-title":"Comparative study of Caffe, Neon, Theano, and Torch for deep learning","author":"bahrampour","year":"2015","journal-title":"ArXiv Preprint"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2015.7404790"},{"key":"ref21","article-title":"The Kaldi speech recognition toolkit","author":"povey","year":"0","journal-title":"IEEE 2011 Workshop on Automatic Speech Recognition and Understanding Rue Marconi 19 Martigny Dec 2011 number Idiap-RR-04-2012 IEEE Signal Processing Society IEEE Catalog No CFP11SRW-USB"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1109\/ICFHR.2016.0052"},{"key":"ref23","first-page":"2000","article-title":"Hierarchical data format version 5","year":"0","journal-title":"The HDF Group"},{"key":"ref26","article-title":"Associative long short-term memory","volume":"abs 1602 3032","author":"danihelka","year":"2016","journal-title":"CoRR"},{"key":"ref25","article-title":"Gated feedback recurrent neural networks","volume":"abs 1502 2367","author":"chung","year":"2015","journal-title":"CoRR"}],"event":{"name":"2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","location":"New Orleans, LA","start":{"date-parts":[[2017,3,5]]},"end":{"date-parts":[[2017,3,9]]}},"container-title":["2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/7943262\/7951776\/07953177.pdf?arnumber=7953177","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2020,10,10]],"date-time":"2020-10-10T10:17:26Z","timestamp":1602325046000},"score":1,"resource":{"primary":{"URL":"http:\/\/ieeexplore.ieee.org\/document\/7953177\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2017,3]]},"references-count":36,"URL":"https:\/\/doi.org\/10.1109\/icassp.2017.7953177","relation":{},"subject":[],"published":{"date-parts":[[2017,3]]}}}