Speech recognition using neural networks - Chapter 9 pptx
Speech recognition using neural networks - Chapter 9 pptx
... 81– 89, 94 , 14 7-1 48 basic operation 81–82 training & testing procedures 8 2-8 4 experiments 8 4-8 7 extensions 8 9- 9 4 weaknesses 9 4 -9 9 vs. HMM 88– 89, 14 7-1 48 LVQ 33, 36, 39, 54, 75, 8 8-8 9, 14 7-1 48 M maximum ... Conference on Neural Networks, IEEE. [78] Lippmann, R. ( 198 9). Review of Neural Networks for Speech Recognition. Neural Computation...
Ngày tải lên: 13/08/2014, 02:21
Speech recognition using neural networks - Chapter 5 doc
... was developed in conjunction with the Janus Speech- to -Speech Translation system at CMU (Waibel et al 199 1, Osterholtz et al 199 2, Woszczyna et al 199 4). While a full discussion of Janus is beyond ... The speech recognition module, for exam- ple, was originally implemented by our LPNN, described in Chapter 6 (Waibel et al 199 1, Osterholtz et al 199 2); but it was later replaced b...
Ngày tải lên: 13/08/2014, 02:21
Speech recognition using neural networks - Chapter 6 pps
... (100%) 2 29/ 2 29 (100%) 3 50/50 (100%) 20/20 (100%) 2 29/ 2 29 (100%) 92 4 1 106/118 (90 %) 55/60 (92 %) 855 /90 0 (95 %) 2 116/118 (98 %) 58/60 (97 %) 886 /90 0 (98 %) 3 117/118 (99 %) 60/60 (100%) 891 /90 0 (99 %) Table ... consid- ered correctly recognized if it appears among the best K candidates. Vocab size Rank Testing set Training set Homophones Novel words 234 1 47/50 (94 %) 19/ 2...
Ngày tải lên: 13/08/2014, 02:21
Speech recognition using neural networks - Chapter 7 pdf
... labeling. Figure 7.14: A 3-state phoneme model outperforms a 1-state phoneme model. 80 82 84 86 88 90 92 94 96 98 100 word accuracy (%) 0 1 2 3 4 5 epochs 1-state vs 3-state models 1 state per ... training set. This important fact has been proven by Gish ( 199 0), Bourlard & Wellekens ( 199 0), Hampshire & Pearlmutter ( 199 0), Ney ( 199 1), and others; see Appendix B for detail...
Ngày tải lên: 13/08/2014, 02:21
Speech recognition using neural networks - Chapter 8 potx
... 402(b) 111 HMM-1 55% HMM-5 96 % 71% 58% 76% HMM-10 97 % 75% 66% 82% LPNN 97 % 60% 41% HCNN 75% LVQ 98 % 84% 74% 61% 83% TDNN 98 % 78% 72% 64% MS-TDNN 98 % 82% 81% 70% 85% Table 8.1: Comparative results ... testing criteria. It is with the MS-TDNN that we achieved a word recognition accuracy of 90 .5% using only 67K parameters, significantly outperforming the context inde- pendent HM...
Ngày tải lên: 13/08/2014, 02:21
speech recognition using neural networks
... Delay Neural Network (TDNN), shown in Figure 3.8. This architecture was initially developed for phoneme recognition (Lang 198 9, Waibel et al 198 9), but it has also been applied to hand- writing recognition ... performance. We will see that neural networks help to avoid this problem. 1.2. Neural Networks Connectionism, or the study of artificial neural networks, was initiall...
Ngày tải lên: 28/04/2014, 10:18