標(biāo)題: Titlebook: Advances in Speech and Language Technologies for Iberian Languages; IberSPEECH 2012 Conf Doroteo Torre Toledano,Alfonso Ortega Giménez ,Dan [打印本頁(yè)] 作者: dilate 時(shí)間: 2025-3-21 16:49
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages影響因子(影響力)
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages影響因子(影響力)學(xué)科排名
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages網(wǎng)絡(luò)公開(kāi)度
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages網(wǎng)絡(luò)公開(kāi)度學(xué)科排名
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages被引頻次
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages被引頻次學(xué)科排名
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages年度引用
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages年度引用學(xué)科排名
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages讀者反饋
書(shū)目名稱Advances in Speech and Language Technologies for Iberian Languages讀者反饋學(xué)科排名
作者: galley 時(shí)間: 2025-3-21 23:47
On the use of Total Variability and Probabilistic Linear Discriminant Analysis for Speaker Verificatverification on short utterances. While the recent advances in the field dealing with the session variability problem have proved to greatly outperform speaker verification systems on typical scenarios where a reasonable amount of speech is available, this performance rapidly degrades at the presenc作者: 品牌 時(shí)間: 2025-3-22 01:57 作者: Bumptious 時(shí)間: 2025-3-22 07:59 作者: Scintigraphy 時(shí)間: 2025-3-22 10:37
Evaluation of a New Beam-Search Formant Tracking Algorithm in Noisy Environments with three formant tracking methods. The proposed formant tracking algorithm makes use of the roots of the polynomial of a Linear Predictive Coding (LPC) as formant candidates. The best combination of formant candidates respect to a defined cost function are selected applying a beam-search algorith作者: Aerophagia 時(shí)間: 2025-3-22 14:07 作者: 決定性 時(shí)間: 2025-3-22 20:28
Preliminary Results of Alignment of Text and Audio in News and Songss. For this purpose two methods are used. The first one is basically a forced alignment process of the audio and text based on pre-existent models. The second one is a model-free method in which new models are trained on the audio to align producing as a result the aligned text and audio. For analys作者: 細(xì)微差別 時(shí)間: 2025-3-22 21:13
Aligning Very Long Speech Signals to Bilingual Transcriptions of Parliamentary Sessionsanscriptions, even when two different languages are employed. The alignment algorithm operates on two phonetic sequences, the first one automatically extracted from the speech signal by means of a phone decoder, and the second one obtained from the reference text by means of a multilingual grapheme-作者: Brain-Imaging 時(shí)間: 2025-3-23 05:28
Factor Analysis Segmentation and Classification in Broadcast News Domain where every input sequence is a language. Following this idea, a study between the classic segmentation systems based on HMM/GMM and FA is done over the output of a perfect segmentation system (oracle boundaries). It can be seen how FA improves the classification results compared to HMM/GMM. Also, 作者: GRIN 時(shí)間: 2025-3-23 07:27
Prosodic and Phonetic Features for Speaking Styles Classification and Detection the segmentation of multimedia data into consistent parts and has important applications, such as identifying speech segments to train acoustic models for speech recognition. In this work the database consists of daily news broadcasts in Portuguese television, on which two main speaking styles are 作者: 監(jiān)禁 時(shí)間: 2025-3-23 13:19
Voice Pathology Detection on the Saarbrücken Voice Database with Calibration and Fusion of Scores Usa discriminative calibration and fusion. The SVD is freely available online containing a collection of voice recordings of different pathologies, including both functional and organic. A generative Gaussian mixture model trained with mel-frequency cepstral coefficients, harmonics-to-noise ratio, nor作者: 緯度 時(shí)間: 2025-3-23 17:11
Score Level versus Audio Level Fusion for Voice Pathology Detection on the Saarbrücken Voice Database containing a collection of voice recordings of different pathologies, both functional and organic. It includes recordings for more than 2000 speakers in which sustained vowels /a/, /i/, and /u/ are pronounced with normal, low, high, and low-high-low intonations. This variety of sounds makes possib作者: Chandelier 時(shí)間: 2025-3-23 18:46 作者: NOTCH 時(shí)間: 2025-3-23 22:32 作者: 中國(guó)紀(jì)念碑 時(shí)間: 2025-3-24 02:58 作者: stratum-corneum 時(shí)間: 2025-3-24 06:45
Mutual Information and Perplexity Based Clustering of Dialogue Information for Dynamic Adaptation ofgue system. The purpose is to estimate a language model related to each cluster, and use them to dynamically modify the model of the speech recognizer at each dialogue turn. In the first approach we build the cluster tree using local decisions based on a Maximum Normalized Mutual Information criteri作者: Celiac-Plexus 時(shí)間: 2025-3-24 11:51 作者: 招致 時(shí)間: 2025-3-24 18:42 作者: altruism 時(shí)間: 2025-3-24 19:13 作者: META 時(shí)間: 2025-3-25 02:01 作者: expansive 時(shí)間: 2025-3-25 06:26
Computer and Information Sciencethe output of a perfect segmentation system (oracle boundaries). It can be seen how FA improves the classification results compared to HMM/GMM. Also, the first experiments of an on-building FA segmentation system are reported suggesting the need to improve the channel compensation over some classes.作者: chandel 時(shí)間: 2025-3-25 08:09 作者: 上漲 時(shí)間: 2025-3-25 11:47
Factor Analysis Segmentation and Classification in Broadcast News Domainthe output of a perfect segmentation system (oracle boundaries). It can be seen how FA improves the classification results compared to HMM/GMM. Also, the first experiments of an on-building FA segmentation system are reported suggesting the need to improve the channel compensation over some classes.作者: 變色龍 時(shí)間: 2025-3-25 16:36
Merging Intention and Emotion to Develop Adaptive Dialogue Systemsived as an intermediate phase between natural language understanding and dialogue management in the architecture of these systems. We have applied and evaluated our method in the UAH system, for which the evaluation results show that merging both sources of information improves system performance as well as its perceived quality.作者: ALE 時(shí)間: 2025-3-25 23:32 作者: Visual-Field 時(shí)間: 2025-3-26 00:29 作者: 駁船 時(shí)間: 2025-3-26 05:13
On the Influence of Automatic Segmentation and Clustering in Automatic Speech Recognitionranscrigal, and results show that the speaker diarization system presented in this work is suitable as a previous step to ASR, as the performance is almost the same as the obtained when using manual segmentation and clustering.作者: 青少年 時(shí)間: 2025-3-26 09:30 作者: PLE 時(shí)間: 2025-3-26 14:09 作者: mortgage 時(shí)間: 2025-3-26 19:42
A Multilingual SLU System Based on Semantic Decoding of Graphs of Wordsed. The graph of words generated in this phase is the input to the semantic decoding algorithm specifically designed to combine statistical models and graphs of words. Some experiments that show the good behavior of the proposed approach are also presented.作者: adroit 時(shí)間: 2025-3-27 00:08 作者: RECUR 時(shí)間: 2025-3-27 02:20 作者: 他姓手中拿著 時(shí)間: 2025-3-27 07:11
The Stochastic Network Calculus Methodology,on. In the second one we take global decisions, based on the optimization of the global perplexity of the combination of the cluster-related LMs. Our experiments show a relative reduction of the word error rate of 15.17%, which helps to improve the performance of the understanding and the dialogue manager modules.作者: 誹謗 時(shí)間: 2025-3-27 13:10
Improving the Quality of Standard GMM-Based Voice Conversion Systems by Considering Physically Motivents in the average quality of the converted speech with respect to traditional statistical methods. This is achieved without modifying the input/output parameters or the shape of the conversion function.作者: dysphagia 時(shí)間: 2025-3-27 15:49 作者: 破布 時(shí)間: 2025-3-27 20:05
1865-0929 erization and recognition; audio and speech segmentation; pathology detection and speech characterization; dialogue and multimodal systems; robustness in automatic speech recognition; applications of speech and language technologies.978-3-642-35291-1978-3-642-35292-8Series ISSN 1865-0929 Series E-ISSN 1865-0937 作者: Crayon 時(shí)間: 2025-3-27 22:44
Jinhyuck Choi,Kwangmi Ko Kim,Yanggon Kimacceptable actual DCF on the degraded dataset. We show how this method can be used to reduce the actual DCF to values lower than 1. We compare results using different quality measures and Bayesian network configurations.作者: Coordinate 時(shí)間: 2025-3-28 05:11
Wentian Ji,Qingju Guo,Yanrui Leiuency of formants. Experiments were carried out with a subset of the TIMIT database, contaminated with various types and levels of noises. The results show that the beam-search formant tracker have a robust behavior in noisy environments and it is clearly more precise than the rest of compared methods.作者: Plaque 時(shí)間: 2025-3-28 08:58
Alfian Akbar Gozali,Shigeru Fujimuraranscrigal, and results show that the speaker diarization system presented in this work is suitable as a previous step to ASR, as the performance is almost the same as the obtained when using manual segmentation and clustering.作者: Incompetent 時(shí)間: 2025-3-28 12:27 作者: MIR 時(shí)間: 2025-3-28 14:55 作者: 刪除 時(shí)間: 2025-3-28 19:02 作者: sigmoid-colon 時(shí)間: 2025-3-29 01:09
Conference proceedings 2012ian SLTech Workshop, held in Madrid, Spain, in November 21-23, 2012. The 29 revised papers were carefully reviewed and selected from 80 submissions. The papers are organized in topical sections on speaker characterization and recognition; audio and speech segmentation; pathology detection and speech作者: cornucopia 時(shí)間: 2025-3-29 07:06
Siya Bao,Masao Yanagisawa,Nozomu Togawa belonging to the NIST Speaker Recognition Evaluation 2010 (NIST SRE10) and it explores the multiple parameters, which define TV and PLDA in order to give some insight about their relevance in this specific scenario.作者: 灌溉 時(shí)間: 2025-3-29 11:16
Thomas Seemann,Harald Hungenberg2 females and 2 males). Analyzing all the results, we observe that news is better aligned than songs, as expected. The two methods work similarly in both . songs and news, but in the case of songs that include the instrumental part, the model-free method is much better.作者: Dappled 時(shí)間: 2025-3-29 11:42 作者: insurgent 時(shí)間: 2025-3-29 19:14 作者: Visual-Field 時(shí)間: 2025-3-29 19:55 作者: fluffy 時(shí)間: 2025-3-30 01:04
Preliminary Results of Alignment of Text and Audio in News and Songs2 females and 2 males). Analyzing all the results, we observe that news is better aligned than songs, as expected. The two methods work similarly in both . songs and news, but in the case of songs that include the instrumental part, the model-free method is much better.作者: Decibel 時(shí)間: 2025-3-30 06:30
Prosodic and Phonetic Features for Speaking Styles Classification and Detectiontep separates the speech segments from the non-speech audio segments and the second step classifies read versus spontaneous speaking style. The use of phonetic and prosodic features provides alternative information that leads to an improvement of the classification and detection task.作者: 即席演說(shuō) 時(shí)間: 2025-3-30 09:11 作者: crockery 時(shí)間: 2025-3-30 12:42
Jinhyuck Choi,Kwangmi Ko Kim,Yanggon Kimork, we use Bayesian networks to model the relations between the speaker verification score, a set of speech quality measures and the trial reliability. We use this model to detect and discard unreliable trials. We present results on the NIST SRE2010 dataset artificially degraded with different type作者: Thrombolysis 時(shí)間: 2025-3-30 20:10 作者: GRE 時(shí)間: 2025-3-30 21:52 作者: Creditee 時(shí)間: 2025-3-31 03:12 作者: yohimbine 時(shí)間: 2025-3-31 07:47 作者: geriatrician 時(shí)間: 2025-3-31 09:39 作者: Absenteeism 時(shí)間: 2025-3-31 17:20
Thomas Seemann,Harald Hungenbergs. For this purpose two methods are used. The first one is basically a forced alignment process of the audio and text based on pre-existent models. The second one is a model-free method in which new models are trained on the audio to align producing as a result the aligned text and audio. For analys作者: dandruff 時(shí)間: 2025-3-31 20:59 作者: 顛簸下上 時(shí)間: 2025-3-31 21:51
