566 工程科学学报，第42卷，第5期 based on the theory and practice of multimedia learning and [20]Szegedy C,loffe S,Vanhoucke V,et al.Inception-v4,Inception- thetoric.Biochem Mol Biol Educ,2017,45(3):249 ResNet and the impact of residual connections on leaming / [2] Welch K E.Electric Rhetoric:Classical Rhetoric,Oralism,and A Proceedings of Thirty-First AAAI Conference on Artificial New Literacy.Cambridge:MIT Press,1999 Intelligence.San Franciso,2017:4278 [3] Berlin James A.Contemporary composition:the major [21]Deng J,Dong W,Socher R,et al.ImageNet:A large-scale pedagogical theories.College English,1982,44(8):765 hierarchical image database /l Proceedings of IEEE Conference [4] O'Halloran K L.Interdependence,interaction and metaphor in on Computer Vision and Pattern Recognition (CVPR 2009) multi-semiotic texts.Social Semiorics,1999,9(3):317 Miami,,2009:248 [5] O'Halloran K L.Classroom discourse in mathematics:a multi- [22]Lample G,Ballesteros M,Subramanian S,et al.Neural semiotic analysis.Linguistics Educ,1998,10(3):359 architectures for named entity recognition /l Proceedings of the [6] Morency L P,Baltrusaitis T.Tutorial on multimodal machine 2016 Conference of the North American Chapter of the learning [R/OL]Language Technologies Institute (2016-6-26) Association for Computational Linguistics:Human Language [2019-03-05].https://www.cs.cmu.edu/-morency/MMML- Technologies.San Diego,2016:260 Tutorial-ACL2017.pdf [23]Ngiam J,Khosla A.Kim M,et al.Multimodal deep learning / [7] Plummer B A,Wang L W,Cervantes C M,et al.Flickr30k Proceedings of the 28th International Conference on Machine entities:collecting region-to-phrase correspondences for richer Learning.Bellevue,2011:689 image-to-sentence models /Proceedings of IEEE International [24]Baltrusaitis T,Ahuja C,Morency L P.Multimodal machine Conference on Computer Vision (ICCV 2015).Santiago,2015: leaming:a survey and taxonomy.IEEE Trans Pattern Anal 2641 Machine Intelligence,2019,41(2):423 [8] von Glasersfeld E,Pisani PP.The multistore parser for [25]Zhang L,Zhao Y,Zhu Z F,et al.Multi-view missing data hierarchical syntactic structures.Commn ACM,1970,13(2):74 completion.IEEE Trans Knowledge Data Eng,2018,30(7): [9] Jackson P.Introduction to Expert Systems.3rd Ed.Boston: 1296 Addison Wesley,1998 [26] Wang L Q,Sun W C,Zhao Z C,et al.Modeling intra-and inter- [10] Cortes C,Vapnik V.Support-vector networks.Machine pair correlation via heterogeneous high-order preserving for Learning,1995,20(3:273 cross-modal retrieval.Signal Process,2017,131:249 [11] Pearl J.Probabilistic Reasoning in Intelligent Systems:Networks [27] Liu H P,Li F X,Xu X Y,et al.Multi-modal local receptive field of Plausible Inference.San Francisco:Morgan Kaufmann extreme learning machine for object recognition.Neurocom Publishers,1988 puting,2018.277:4 [12]Jelinek F.Statistical Methods for Speech Recognition [28]Fu K,Jin J Q,Cui R P,et al.Aligning where to see and what to Cambridge:MIT Press,1997 tell:image captioning with region-based attention and scene [13]McGurk H,MacDonald J.Hearing lips and seeing voices specific contexts.IEEE Trans Pattern Anal Machine Intelligence Nature,.1976,264(5588):746 2017,39(12):2321 [14]Petajan E D.Automatic Lipreading to Enhance Speech [29]Breiman L.Random forests.Machine Learning,2001,45(1):5 Recognition (Speech Reading)[Dissertation].University of [30]Breiman L,Friedman J H,Olshen R A,et al.Classification and Illinois at Urbana-Champaign,1984 Regression Trees.Florida:Chapman and Hall/CRC,1998 [15]Fels S S,Hinton G E.Glove-Talk:a neural network interface [31]Breiman L.Statistical modeling:the two cultures.Statist Sci, between a data-glove and a speech synthesizer.IEEE Trans 2001.16(3):199 Neural Nenorks,1993,4(1):2 [32]Vapnik V N,Cervonenkis,A.J.Empirical Inference.Berlin: [16]Srivastava N,Hinton G,Krizhevsky A,et al.Dropout:a simple Springer,2013 way to prevent neural networks from overfitting.J Machine [33]Scholkopf B,Smola A J,Bach F.Learning with Kernels:Support Learning Res,2014,15:1929 Vector Machines,Regularication,Optimication,and Beyond. [17]Glorot X,Bordes A.Bengio Y.Deep sparse rectifier neural Cambridge:MIT Press,2002 networks /l Proceedings of the 14th International Conference on [34]Mercer J.Functions of positive and negative type,and their Artificial Intelligence and Statistics.Fort Lauderdale,2011:315 connection with the theory of integral equations.Philos Trans R [18]He K M,Zhang X Y,Ren S Q,et al.Deep residual leaming for Soc London Ser4,1909,209(441-458:415 image recognition /Proceedings of IEEE Conference on [35]Aronszain N.Theory of reproducing kernels.Trans Am Math Computer Vision and Pattern Recognition(CVPR 2016).Las Soc,1950,68(3):337 Vegas,2016:770 [36]Steinwart I,Hush D,Scovel C.An explicit description of the [19]Krizhevsky A,Sutskever I,Hinton G E.ImageNet classification reproducing kemel Hilbert spaces of Gaussian RBF kernels. with deep convolutional neural networks /Advances in Neural JEEE Trans Inf Theory,2006,52(10):4635 Information Processing Systems.Lake Tahoe,2012:1097 [37]Lodhi H,Saunders C,Shawe-Taylor J,et al.Text classificationbased  on  the  theory  and  practice  of  multimedia  learning  and rhetoric. Biochem Mol Biol Educ, 2017, 45（3）: 249 Welch K E. Electric Rhetoric: Classical Rhetoric, Oralism, and A New Literacy. Cambridge: MIT Press, 1999 [2] Berlin  James  A.  Contemporary  composition:  the  major pedagogical theories. College English, 1982, 44（8）: 765 [3] O'Halloran  K  L.  Interdependence,  interaction  and  metaphor  in multi-semiotic texts. Social Semiotics, 1999, 9（3）: 317 [4] O'Halloran  K  L.  Classroom  discourse  in  mathematics:  a  multi￾semiotic analysis. Linguistics Educ, 1998, 10（3）: 359 [5] Morency  L  P,  Baltrusaitis  T.  Tutorial  on  multimodal  machine learning  [R/OL]. Language Technologies Institute (2016-6-26) [2019-03-05]. https://www.cs.cmu.edu/~morency/MMML￾Tutorial-ACL2017.pdf [6] Plummer  B  A,  Wang  L  W,  Cervantes  C  M,  et  al.  Flickr30k entities:  collecting  region-to-phrase  correspondences  for  richer image-to-sentence  models  // Proceedings of IEEE International Conference on Computer Vision (ICCV 2015).  Santiago,  2015: 2641 [7] von  Glasersfeld  E,  Pisani  P  P.  The  multistore  parser  for hierarchical syntactic structures. Commun ACM, 1970, 13（2）: 74 [8] Jackson  P. Introduction to Expert Systems.  3rd  Ed.  Boston: Addison Wesley, 1998 [9] Cortes  C,  Vapnik  V.  Support-vector  networks. Machine Learning, 1995, 20（3）: 273 [10] Pearl J. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference.  San  Francisco:  Morgan  Kaufmann Publishers, 1988 [11] Jelinek  F. Statistical Methods for Speech Recognition. Cambridge: MIT Press, 1997 [12] McGurk  H,  MacDonald  J.  Hearing  lips  and  seeing  voices. Nature, 1976, 264（5588）: 746 [13] Petajan  E  D. Automatic Lipreading to Enhance Speech Recognition (Speech Reading)  [Dissertation].  University  of Illinois at Urbana-Champaign, 1984 [14] Fels  S  S,  Hinton  G  E.  Glove-Talk:  a  neural  network  interface between  a  data-glove  and  a  speech  synthesizer. IEEE Trans Neural Networks, 1993, 4（1）: 2 [15] Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: a simple way  to  prevent  neural  networks  from  overfitting. J Machine Learning Res, 2014, 15: 1929 [16] Glorot  X,  Bordes  A,  Bengio  Y.  Deep  sparse  rectifier  neural networks // Proceedings of the 14th International Conference on Artificial Intelligence and Statistics. Fort Lauderdale, 2011: 315 [17] He K M, Zhang X Y, Ren S Q, et al. Deep residual learning for image  recognition  // Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016).  Las Vegas, 2016: 770 [18] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with  deep  convolutional  neural  networks  // Advances in Neural Information Processing Systems. Lake Tahoe, 2012: 1097 [19] Szegedy C, Ioffe S, Vanhoucke V, et al. Inception-v4, Inception￾ResNet  and  the  impact  of  residual  connections  on  learning  // Proceedings of Thirty-First AAAI Conference on Artificial Intelligence. San Franciso, 2017: 4278 [20] Deng  J,  Dong  W,  Socher  R,  et  al.  ImageNet:  A  large-scale hierarchical  image  database  // Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2009). Miami, 2009: 248 [21] Lample  G,  Ballesteros  M,  Subramanian  S,  et  al.  Neural architectures  for  named  entity  recognition  // Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. San Diego, 2016: 260 [22] Ngiam  J,  Khosla  A,  Kim  M,  et  al.  Multimodal  deep  learning  // Proceedings of the 28th International Conference on Machine Learning. Bellevue, 2011: 689 [23] Baltrusaitis  T,  Ahuja  C,  Morency  L  P.  Multimodal  machine learning:  a  survey  and  taxonomy. IEEE Trans Pattern Anal Machine Intelligence, 2019, 41（2）: 423 [24] Zhang  L,  Zhao  Y,  Zhu  Z  F,  et  al.  Multi-view  missing  data completion. IEEE Trans Knowledge Data Eng,  2018,  30（7）: 1296 [25] Wang L Q, Sun W C, Zhao Z C, et al. Modeling intra- and inter￾pair  correlation  via  heterogeneous  high-order  preserving  for cross-modal retrieval. Signal Process, 2017, 131: 249 [26] Liu H P, Li F X, Xu X Y, et al. Multi-modal local receptive field extreme  learning  machine  for  object  recognition. Neurocom￾puting, 2018, 277: 4 [27] Fu K, Jin J Q, Cui R P, et al. Aligning where to see and what to tell:  image  captioning  with  region-based  attention  and  scene￾specific contexts. IEEE Trans Pattern Anal Machine Intelligence, 2017, 39（12）: 2321 [28] [29] Breiman L. Random forests. Machine Learning, 2001, 45（1）: 5 Breiman L, Friedman J H, Olshen R A, et al. Classification and Regression Trees. Florida: Chapman and Hall/CRC, 1998 [30] Breiman  L.  Statistical  modeling:  the  two  cultures. Statist Sci, 2001, 16（3）: 199 [31] Vapnik  V  N,  Cervonenkis,  A.  J. Empirical Inference. Berlin: Springer, 2013 [32] Schölkopf B, Smola A J, Bach F. Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. Cambridge: MIT Press, 2002 [33] Mercer  J.  Functions  of  positive  and  negative  type,  and  their connection with the theory of integral equations. Philos Trans R Soc London Ser A, 1909, 209（441-458）: 415 [34] Aronszajn  N.  Theory  of  reproducing  kernels. Trans Am Math Soc, 1950, 68（3）: 337 [35] Steinwart  I,  Hush  D,  Scovel  C.  An  explicit  description  of  the reproducing  kernel  Hilbert  spaces  of  Gaussian  RBF  kernels. IEEE Trans Inf Theory, 2006, 52（10）: 4635 [36] [37] Lodhi  H,  Saunders  C,  Shawe-Taylor  J,  et  al.  Text  classification · 566 · 工程科学学报，第 42 卷，第 5 期