000 03231nam a22004455i 4500
001 978-3-642-24797-2
003 DE-He213
005 20140220083304.0
007 cr nn 008mamaa
008 120205s2012 gw | s |||| 0|eng d
020 _a9783642247972
_9978-3-642-24797-2
024 7 _a10.1007/978-3-642-24797-2
_2doi
050 4 _aQ342
072 7 _aUYQ
_2bicssc
072 7 _aCOM004000
_2bisacsh
082 0 4 _a006.3
_223
100 1 _aGraves, Alex.
_eauthor.
245 1 0 _aSupervised Sequence Labelling with Recurrent Neural Networks
_h[electronic resource] /
_cby Alex Graves.
264 1 _aBerlin, Heidelberg :
_bSpringer Berlin Heidelberg,
_c2012.
300 _aXIV, 146p. 50 illus., 12 illus. in color.
_bonline resource.
336 _atext
_btxt
_2rdacontent
337 _acomputer
_bc
_2rdamedia
338 _aonline resource
_bcr
_2rdacarrier
347 _atext file
_bPDF
_2rda
490 1 _aStudies in Computational Intelligence,
_x1860-949X ;
_v385
505 0 _aIntroduction -- Supervised Sequence Labelling -- Neural Networks -- Long Short-Term Memory -- A Comparison of Network Architectures -- Hidden Markov Model Hybrids -- Connectionist Temporal Classification -- Multidimensional Networks -- Hierarchical Subsampling Networks.
520 _aSupervised sequence labelling is a vital area of machine learning, encompassing tasks such as speech, handwriting and gesture recognition, protein secondary structure prediction and part-of-speech tagging. Recurrent neural networks are powerful sequence learning tools—robust to input noise and distortion, able to exploit long-range contextual information—that would seem ideally suited to such problems. However their role in large-scale sequence labelling systems has so far been auxiliary.    The goal of this book is a complete framework for classifying and transcribing sequential data with recurrent neural networks only. Three main innovations are introduced in order to realise this goal. Firstly, the connectionist temporal classification output layer allows the framework to be trained with unsegmented target sequences, such as phoneme-level speech transcriptions; this is in contrast to previous connectionist approaches, which were dependent on error-prone prior segmentation. Secondly, multidimensional recurrent neural networks extend the framework in a natural way to data with more than one spatio-temporal dimension, such as images and videos. Thirdly, the use of hierarchical subsampling makes it feasible to apply the framework to very large or high resolution sequences, such as raw audio or video.   Experimental validation is provided by state-of-the-art results in speech and handwriting recognition.
650 0 _aEngineering.
650 0 _aArtificial intelligence.
650 1 4 _aEngineering.
650 2 4 _aComputational Intelligence.
650 2 4 _aArtificial Intelligence (incl. Robotics).
710 2 _aSpringerLink (Online service)
773 0 _tSpringer eBooks
776 0 8 _iPrinted edition:
_z9783642247965
830 0 _aStudies in Computational Intelligence,
_x1860-949X ;
_v385
856 4 0 _uhttp://dx.doi.org/10.1007/978-3-642-24797-2
912 _aZDB-2-ENG
999 _c102343
_d102343