INTERSPEECH 2010

Teaser Browsing | Click Here

Title Browsing | Click Here

Wall of Fame | Most Viewed INTERSPEECH-2010 Paper

click to vote

INTERSPEECH
2010

242views Signal Processing» more INTERSPEECH 2010»

Recurrent neural network based language model

13 years 6 months ago

Download www.fit.vutbr.cz

A new recurrent neural network based language model (RNN LM) with applications to speech recognition is presented. Results indicate that it is possible to obtain around 50% reduct...

Tomas Mikolov, Martin Karafiát, Lukas Burge...

claim paper

Read More »

Disclaimer and Copyright Notice

Sciweavers respects the rights of all copyright holders and in this regard, authors are only allowed to share a link to their preprint paper on their own website. Every contribution is associated with a desciptive image. It is the sole responsibility of the authors to ensure that their posted image is not copyright infringing. This service is compliant with IEEE copyright.

Id	Views	Title	Status

1	242	Recurrent neural network based language model	claim
2	238	Excitation modeling based on waveform interpolation for HMM-based speech synthesis	claim
3	224	Speaker and language adaptive training for HMM-based polyglot speech synthesis	claim
4	220	Boosted mixture learning of Gaussian mixture HMMs for speech recognition	claim
5	215	Bayesian speaker recognition using Gaussian mixture model and laplace approximation	claim
6	203	Extractive summarization using a latent variable model	claim
7	200	Towards a robust face recognition system using compressive sensing	claim
8	197	Prosodic speaker verification using subspace multinomial models with intersession compensation	claim
9	196	Hierarchical bottle neck features for LVCSR	claim
10	194	Revisiting VTLN using linear transformation on conventional MFCC	claim
11	189	Efficient combined approach for named entity recognition in spoken language	claim
12	185	Adaptation of a tongue shape model by local feature transformations	claim
13	181	Channel detectors for system fusion in the context of NIST LRE 2009	claim
14	181	Active appearance models for photorealistic visual speech synthesis	claim
15	181	Binary coding of speech spectrograms using a deep auto-encoder	claim
16	180	Deep-structured hidden conditional random fields for phonetic recognition	claim
17	180	Phone mismatch penalty matrices for two-stage keyword spotting via multi-pass phone recognizer	claim
18	178	Probabilistic integration of joint density model and speaker model for voice conversion	claim
19	175	Still talking to machines (cognitively speaking)	claim
20	171	Text normalization based on statistical machine translation and internet user support	claim
21	170	Coping imbalanced prosodic unit boundary detection with linguistically-motivated prosodic features	claim
22	170	SCARF: a segmental conditional random field toolkit for speech recognition	claim
23	167	On-the-fly lattice rescoring for real-time automatic speech recognition	claim
24	167	Canonical state models for automatic speech recognition	claim
25	166	Emotion recognition using imperfect speech recognition	claim
26	165	Memory-based active learning for French broadcast news	claim
27	163	A new binary mask based on noise constraints for improved speech intelligibility	claim
28	162	Semi-supervised training of Gaussian mixture models by conditional entropy minimization	claim
29	158	Combining monaural and binaural evidence for reverberant speech segregation	claim
30	156	The RWTH 2009 quaero ASR evaluation system for English and German	claim
31	156	Brno university of technology system for interspeech 2010 paralinguistic challenge	claim
32	155	VAD-measure-embedded decoder with online model adaptation	claim
33	154	Nucleus position within the intonation phrase: a typological study of English, Czech and Hungarian	claim
34	154	Domain adaptation and compensation for emotion detection	claim
35	154	Beyond sentence prosody	claim
36	153	Acoustic feature diversity and speaker verification	claim
37	153	Voice quality evaluation of recent open source codecs	claim
38	153	Acoustic vector resampling for GMMSVM-based speaker verification	claim
39	152	Combining five acoustic level modeling methods for automatic speaker age and gender recognition	claim
40	151	Can tongue be recovered from face? the answer of data-driven statistical models	claim
41	151	Articulatory synthesis and perception of plosive-vowel syllables with virtual consonant targets	claim
42	151	An implementation of decision tree-based context clustering on graphics processing units	claim
43	151	2010, a speech oddity: phonetic transcription of reversed speech	claim
44	148	Speech dominoes and phonetic convergence	claim
45	148	Mask estimation in non-stationary noise environments for missing feature based robust speech recognition	claim
46	148	Voice attributes affecting likability perception	claim
47	147	Sparse component analysis for speech recognition in multi-speaker environment	claim
48	146	SAFE: a statistical algorithm for F0 estimation for both clean and noisy speech	claim
49	145	Data-dependent evaluator modeling and its application to emotional valence classification from speech	claim
50	144	Regularized-MLLR speaker adaptation for computer-assisted language learning system	claim
51	144	Hidden Markov models with context-sensitive observations for grapheme-to-phoneme conversion	claim
52	143	Automatic derivation of phonological rules for mispronunciation detection in a computer-assisted pronunciation training system	claim
53	143	Kinematic analysis of tongue movement control in spastic dysarthria	claim
54	143	HMM-based automatic visual speech segmentation using facial data	claim
55	142	Automatic estimation of transcription accuracy and difficulty	claim
56	142	The relevance of timing, pauses and overlaps in dialogues: detecting topic changes in scenario based meetings	claim
57	142	Exploring recognition network representations for efficient speech inference on highly parallel platforms	claim
58	141	The influence of actual and perceived sexual orientation on diadochokinetic rate in women and men	claim
59	140	Automatic speech recognition system channel modeling	claim
60	140	Augmented set of features for confidence estimation in spoken term detection	claim
61	140	An improved wavelet-based dereverberation for robust automatic speech recognition	claim
62	140	Robust automatic speech recognition with decoder oriented ideal binary mask estimation	claim
63	140	Using dependency parsing and machine learning for factoid question answering on spoken documents	claim
64	139	Feature versus model based noise robustness	claim
65	138	Discriminative adaptation for log-linear acoustic models	claim
66	137	Study on interaction between entropy pruning and kneser-ney smoothing	claim
67	137	Decision tree state clustering with word and syllable features	claim
68	137	Modeling liaison in French by using decision trees	claim
69	136	Vocabulary independent spoken query: a case for subword units	claim
70	136	Unsupervised discovery and training of maximally dissimilar cluster models	claim
71	135	Paraphrase generation to improve text-to-speech synthesis	claim
72	135	The INTERSPEECH 2010 paralinguistic challenge	claim
73	135	Automatic speech recognition for assistive writing in speech supplemented word prediction	claim
74	135	Validation of a training method for L2 continuous-speech segmentation	claim
75	134	Topic and style-adapted language modeling for Thai broadcast news ASR	claim
76	132	Multichannel noise reduction using low order RTF estimate	claim
77	132	Augmentation of adaptation data	claim
78	132	Improvements to the equal-parameter BIC for speaker diarization	claim
79	131	Asymptotically exact noise-corrupted speech likelihoods	claim
80	130	Combination of probabilistic and possibilistic language models	claim
81	130	HMM adaptation using linear spline interpolation with integrated spline parameter training for robust speech recognition	claim
82	129	Constructing Japanese test collections for spoken term detection	claim
83	129	Exploring speaker characteristics for meeting summarization	claim
84	129	On the potential of glottal signatures for speaker recognition	claim
85	128	A discriminative splitting criterion for phonetic decision trees	claim
86	128	Similarity of effects of emotions on the speech organ configuration with and without speaking	claim
87	128	Multi-pitch estimation by a joint 2-d representation of pitch and pitch dynamics	claim
88	128	Full body aero-tactile integration in speech perception	claim
89	128	Identifying articulatory goals from kinematic data using principal differential analysis	claim
90	128	Using a DBN to integrate sparse classification and GMM-based ASR	claim
91	128	Cross-cultural investigation of prosody in verbal feedback in interactional rapport	claim
92	128	A comparative study of constrained and unconstrained approaches for segmentation of speech signal	claim
93	128	Accurate pitch marking for prosodic modification of speech segments	claim
94	127	Setup for acoustic-visual speech synthesis by concatenating bimodal units	claim
95	127	Combining user intention and error modeling for statistical dialog simulators	claim
96	127	Integration of multilayer regression analysis with structure-based pronunciation assessment	claim
97	126	On the interdependencies between voice quality, glottal gaps, and voice-source related acoustic measures	claim
98	126	Manipulating treacheoesophageal speech	claim
99	126	On speaker adaptive training of artificial neural networks	claim
100	126	A factorial sparse coder model for single channel source separation	claim
101	126	A novel hybrid approach for Mandarin speech synthesis	claim
102	125	What else is new than the hamming window? robust MFCCs for speaker recognition via multitapering	claim
103	125	The relation between pitch perception preference and emotion identification	claim
104	124	FSM-based pronunciation modeling using articulatory phonological code	claim
105	124	Efficient HMM-based estimation of missing features, with applications to packet loss concealment	claim
106	124	Semi-supervised part-of-speech tagging in speech applications	claim
107	124	Phonetic subspace mixture model for speaker diarization	claim
108	123	Investigation of full-sequence training of deep belief networks for speech recognition	claim
109	123	Can conversational word usage be used to predict speaker demographics?	claim
110	123	Measuring basic tempo across languages and some implications for speech rhythm	claim
111	123	A comparative large scale study of MLP features for Mandarin ASR	claim
112	123	A spectral LF model based approach to voice source parameterisation	claim
113	122	Say what? why users choose to speak their web queries	claim
114	122	Building transcribed speech corpora quickly and cheaply for many languages	claim
115	122	Comparison of approaches for instrumentally predicting the quality of text-to-speech systems	claim
116	122	Continuous speech recognition with a TF-IDF acoustic model	claim
117	122	Invariant integration features combined with speaker-adaptation methods	claim
118	122	Robust and efficient pitch estimation using an iterative ARMA technique	claim
119	121	HMM-based prosodic structure model using rich linguistic context	claim
120	121	Learning from human errors: prediction of phoneme confusions based on modified ASR training	claim
121	121	Dynamic model selection for spectral voice conversion	claim
122	120	Automatic selection of thresholds for signal separation algorithms based on interaural delay	claim
123	120	Rapid development of speech translation using consecutive interpretation	claim
124	120	Predicting word accuracy for the automatic speech recognition of non-native speech	claim
125	119	Language model cross adaptation for LVCSR system combination	claim
126	119	Context adaptive training with factorized decision trees for HMM-based speech synthesis	claim
127	118	On the relation of Bayes risk, word error, and word posteriors in ASR	claim
128	118	Combining many alignments for speech to speech translation	claim
129	118	Expectations for discourse genre identification: a prosodic study	claim
130	118	Dialect recognition using a phone-GMM-supervector-based SVM kernel	claim
131	117	Approaching human listener accuracy with modern speaker verification	claim
132	117	Prosody and voice quality of vocal social signals: the case of dominance in scenario meetings	claim
133	117	Improved real-time MRI of oral-velar coordination using a golden-ratio spiral view order	claim
134	117	Rapid bootstrapping of five eastern european languages using the rapid language adaptation toolkit	claim
135	117	Comparison of HMM and TMDN methods for lip synchronisation	claim
136	117	Reducing musical noise in blind source separation by time-domain sparse filters and split bregman method	claim
137	117	Artificial and online acquired noise dictionaries for noise robust ASR	claim
138	116	Text-based unstressed syllable prediction in Mandarin	claim
139	116	Glottal-based analysis of the lombard effect	claim
140	116	Speaking style dependency of formant targets	claim
141	116	Advanced speech communication system for deaf people	claim
142	115	Prominence detection in Swedish using syllable correlates	claim
143	115	Landmark-based automated pronunciation error detection	claim
144	115	The use of subvector quantization and discrete densities for fast GMM computation for speaker verification	claim
145	115	Hierarchical multilayer perceptron based language identification	claim
146	115	Semantic facilitation in bilingual everyday speech comprehension	claim
147	114	Fully automatic segmentation for prosodic speech corpora	claim
148	114	Data pruning for template-based automatic speech recognition	claim
149	114	Convexity and fast speech extraction by split bregman method	claim
150	114	Nonlinear enhancement of onset for robust speech recognition	claim
151	114	Phonetic realization of second occurrence focus in Japanese	claim
152	114	Exploiting glottal formant parameters for glottal inverse filtering and parameterization	claim
153	113	Enhancements of viterbi search for fast unit selection synthesis	claim
154	113	A rule-based backchannel prediction model using pitch and pause information	claim
155	112	Within and across sentence boundary language model	claim
156	112	Investigating articulatory setting - pauses, ready position, and rest - using real-time MRI	claim
157	111	Voice search for development	claim
158	111	Analyzing user utterances in barge-in-able spoken dialogue system for improving identification accuracy	claim
159	110	Non-negative matrix factorization based compensation of music for automatic speech recognition	claim
160	110	Wiktionary as a source for automatic pronunciation extraction	claim
161	110	Reduction of broadband noise in speech signals by multilinear subspace analysis	claim
162	109	Online adaptive learning for speech recognition decoding	claim
163	109	Distribution and trichotomic realization of voiced velars in Japanese - an experimental study	claim
164	109	Chirp complex cepstrum-based decomposition for asynchronous glottal analysis	claim
165	108	Prosodic timing analysis for articulatory re-synthesis using a bank of resonators with an adaptive oscillator	claim
166	108	Incremental word learning using large-margin discriminative training and variance floor estimation	claim
167	107	Estimating missing data sequences in x-ray microbeam recordings	claim
168	107	Towards mixed language speech recognition systems	claim
169	107	Modeling pronunciation variation with context-dependent articulatory feature decision trees	claim
170	106	A spoken term detection framework for recovering out-of-vocabulary words using the web	claim
171	106	Laryngealization and features for Chinese tonal recognition	claim
172	106	Extending the punctuation module for european portuguese	claim
173	105	Learning a language model from continuous speech	claim
174	105	Performance estimation of noisy speech recognition considering recognition task complexity	claim
175	104	Language specific effects of emotion on phoneme duration	claim
176	104	Detecting Politeness and efficiency in a cooperative social interaction	claim
177	104	Improved language recognition using mixture components statistics	claim
178	104	Audio-visual anticipatory coarticulation modeling by human and machine	claim
179	103	Active word learning under uncertain input conditions	claim
180	103	Relying on critical articulators to estimate vocal tract spectra in an articulatory-acoustic database	claim
181	103	Dialog prediction for a general model of turn-taking	claim
182	102	Semi-automated update of automatic transcription system for the Japanese national congress	claim
183	102	Native and non-native speaker judgements on the quality of synthesized speech	claim
184	102	An HMM trajectory tiling (HTT) approach to high quality TTS	claim
185	102	Articulatory-functional modeling of speech prosody: a review	claim
186	101	Unscented transform with online distortion estimation for HMM adaptation	claim
187	101	Evaluation of speaker mimic technology for personalizing SGD voices	claim
188	101	Direct observation of pruning errors (DOPE): a search analysis tool	claim
189	101	Strategies for statistical spoken language understanding with small amount of data - an empirical study	claim
190	101	AutoBI - a tool for automatic toBI annotation	claim
191	101	Impact of lack of acoustic feedback in EMG-based silent speech recognition	claim
192	100	Semi-supervised extractive speech summarization via co-training algorithm	claim
193	100	Improved neural network based language modelling and adaptation	claim
194	99	Roles of the average voice in speaker-adaptive HMM-based speech synthesis	claim
195	99	Multimodal speaker diarization using oriented optical flow histograms	claim
196	99	Towards affective state modeling in narrative and conversational settings	claim
197	99	Acoustic feature analysis in speech emotion primitives estimation	claim
198	99	On generating combilex pronunciations via morphological analysis	claim
199	98	Lexical entrainment of real users in the let's go spoken dialog system	claim
200	98	Competition in the perception of spoken Japanese words	claim
201	98	Synthesis of fast speech with interpolation of adapted HSMMs and its evaluation by blind and sighted listeners	claim
202	96	Shape-invariant speech transformation with the phase vocoder	claim
203	95	Quality-based playout buffering with FEC for conversational voIP	claim
204	95	An exploration of voice source correlates of focus	claim
205	95	Learning speaker normalization using semisupervised manifold alignment	claim
206	95	Improvements of search error risk minimization in viterbi beam search for speech recognition	claim
207	95	A classifier-based target cost for unit selection speech synthesis trained on perceptual data	claim
208	94	Frequency of occurrence effects on pitch accent realisation	claim
209	94	Towards spoken term discovery at scale with zero resources	claim
210	94	The prosody of Swedish conversational grunts	claim
211	94	On using voice source measures in automatic gender classification of children's speech	claim
212	94	HMM-based text-to-articulatory-movement prediction and analysis of critical articulators	claim
213	93	Frequency-domain delexicalization using surrogate vowels	claim
214	93	Automatic discriminative measurement of voice onset time	claim
215	93	Similarity scoring for recognizing repeated out-of-vocabulary words	claim
216	93	Discriminative adaptation based on fast combination of DMAP and dfMLLR	claim
217	93	What do you mean, you're uncertain?: the interpretation of cue words and rising intonation in dialogue	claim
218	93	Detecting categorical perception in continuous discrimination data	claim
219	92	Search by voice in Mandarin Chinese	claim
220	92	Improving monaural speaker identification by double-talk detection	claim
221	91	Rhythm and formant features for automatic alcohol detection	claim
222	90	Exploring web-browser based runtimes engines for creating ubiquitous speech interfaces	claim
223	90	Verifying pronunciation dictionaries using conflict analysis	claim
224	90	Pitch similarity in the vicinity of backchannels	claim
225	88	Exploring the mechanism of tonal contraction in taiwan Mandarin	claim
226	88	A procedure for estimating gestural scores from natural speech	claim
227	88	Predicting human perception and ASR classification of word-final [t] by its acoustic sub-segmental properties	claim
228	86	The impact of ASR on abstractive vs. extractive meeting summaries	claim
229	86	Morphological and predictability effects on schwa reduction: the case of dutch word-initial syllables	claim
230	86	Disambiguating the functions of conversational sounds with prosody: the case of yeah	claim
231	85	Autoregressive clustering for HMM speech synthesis	claim
232	85	An intonation model for TTS in sepedi	claim
233	85	Redescribing intonational categories with functional data analysis	claim
234	85	Influence of musical training on perception of L2 speech	claim
235	84	Korean lenis, fortis, and aspirated stops: effect of place of articulation on acoustic realization	claim
236	84	Specification in context - devoicing processes in Polish, French, american English and German sonorants	claim
237	83	Speech synthesis by modeling harmonics structure with multiple function	claim
238	83	Cross-lingual talker discrimination	claim
239	83	Phrase-medial vowel devoicing in spontaneous French	claim
240	82	Detection of hot spots in poster conversations based on reactive tokens of audience	claim
241	81	The role of higher-level linguistic features in HMM-based speech synthesis	claim
242	80	Minimally invasive surgery for spoken dialog systems	claim
243	77	Direct construction of compact context-dependency transducers from data	claim
244	72	Prominence based scoring of speech segments for automatic speech-to-speech summarization	claim
245	71	Time conditioned search in automatic speech recognition reconsidered	claim
246	56	Influence of lexical tones on intonation in kammu	claim

Sciweavers

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers