| IEICE Technical Committee Submission System Conference Schedule |
Online Proceedings [Sign in] Tech. Rep. Archives |
All Technical Committee Conferences (Searched in: All Years)
| Committee | Date Time | Place | Paper Title / Authors | Abstract | Paper # | |
| SIP, SP, EA, IPSJ-SLP [detail] | 2024-03-01 16:05 |
Okinawa | (Primary: On-site, Secondary: Online) |
Evaluating speech generation based on objective measures for text generation Takaaki Saeki (UTokyo), Soumi Maiti (CMU), Shinnosuke Takamichi (UTokyo), Shinji Watanabe (CMU), Hiroshi Saruwatari (UTokyo)  EA2023-133 SIP2023-180 SP2023-115 |
In the evaluation of speech generation, while subjective judgments have long been the gold standard, objective metrics s... [more] | EA2023-133 SIP2023-180 SP2023-115 pp.421-426 |
| SP, IPSJ-SLP, IPSJ-MUS | 2021-06-19 09:30 |
Online | Online | [Invited Talk]
Toward a Unification of Various Speech Processing Tasks Based on End-to-End Neural networks Shinji Watanabe (CMU) SP2021-8 |
This presentation will introduce the recent progress of speech processing technologies based on end-to-end neural networ... [more] | SP2021-8 p.38 |
| NLC, IPSJ-NL, SP, IPSJ-SLP [detail] | 2020-12-02 09:40 |
Online | Online | Fast End-to-End Speech Recognition with CTC and Mask Predict Yosuke Higuchi (Waseda Univ.), Hirofumi Inaguma (Kyoto Univ.), Shinji Watanabe (JHU), Tetsuji Ogawa, Tetsunori Kobayashi (Waseda Univ.) NLC2020-13 SP2020-16 |
We present a fast non-autoregressive (NAR) end-to-end automatic speech recognition (E2E-ASR) framework, which generates ... [more] | NLC2020-13 SP2020-16 pp.1-6 |
| EA, ASJ-H | 2017-07-20 13:40 |
Hokkaido | Hokkaido Univ. | Explicit Event Duration-Controlled BLSTM-HSMM Hybrid Model for Polyphonic Sound Event Detection Tomoki Hayashi (Nagoya Univ.), Shinji Watanabe (MERL), Tomoki Toda (Nagoya Univ.), Takaaki Hori, JonathanLe Roux (MERL), Kazuya Takeda (Nagoya Univ.) EA2017-2 |
This paper presents a new BLSTM-HSMM hybrid approach for polyphonic Sound Event Detection (SED). It builds upon a state-... [more] | EA2017-2 pp.9-14 |
| SP, IPSJ-SLP (Joint) |
2016-07-30 09:00 |
Yamagata | Takinoyu Hotel | Discriminative Training Method of Recurrent Neural Network Language Models for Speech Recognition Yuuki Tachioka (Mitsubishi Electric), Shinji Watanabe (MERL) SP2016-26 |
A recurrent neural network language model (RNN-LM) can consider a longer word context than an n-gram language model, and... [more] | SP2016-26 pp.33-38 |
| NLC, IPSJ-NL, SP, IPSJ-SLP (Joint) [detail] |
2015-12-02 13:55 |
Aichi | Nagoya Inst of Tech. | Automation of high performance system building for large vocabulary speech recognition using evolution strategy with pareto optimality Takafumi Moriya, Tomohiro Tanaka, Takahiro Shinozaki (Tokyo Tech), Shinji Watanabe (MERL), Kevin Duh (NAIST) SP2015-75 |
The performance of speech recognition tasks can be significantly improved by the use of deep neural networks (DNN). Howe... [more] | SP2015-75 pp.31-36 |
| SP, IPSJ-SLP (Joint) |
2015-07-16 16:20 |
Nagano | Katakura Suwako Hotel | Sequence Discriminative Training for Low-Rank Deep Neural Networks Yuuki Tachioka (Mitsubishi Electric), Shinji Watanabe, Jonathan Le Roux, John Hershey (MERL) SP2015-39 |
Deep neural network (DNN) acoustic models outperform conventional Gaussian mixture model (GMM) but the number of paramet... [more] | SP2015-39 pp.19-24 |
| SP, IPSJ-SLP (Joint) |
2015-07-16 16:50 |
Nagano | Katakura Suwako Hotel | A Feature-Space Adaptation Technique using Regression Tree-based Multiple Transformation Matrices Hiroki Kanagawa, Yuuki Tachioka (Mitsubishi Electric Corp.), Shinji Watanabe (MERL), Jun Ishii (Mitsubishi Electric Corp.) SP2015-40 |
(To be available after the conference date) [more] | SP2015-40 pp.25-30 |
| SP, IPSJ-SLP (Joint) |
2014-07-25 09:30 |
Iwate | Hotel Hanamaki | A generalized discriminative training framework for system combination Yuuki Tachioka, Shinji Watanabe, Jonathan Le Roux, John Hershey (Mitsubishi Electric) SP2014-65 |
This paper proposes a generalized discriminative training framework for system combination, which encompasses acoustic m... [more] | SP2014-65 pp.13-18 |
| SP, IPSJ-MUS | 2014-05-24 11:30 |
Tokyo | Discriminative training of acoustic models for system combination Yuuki Tachioka (Mitsubishi Electric), Shinji Watanabe, Jonathan Le Roux, John R. Hershey (MERL) SP2014-15 |
In discriminative training methods, the objective function is designed to improve the performance of automatic speech re... [more] | SP2014-15 pp.147-152 |
|
| SP, IPSJ-SLP (Joint) |
2013-07-25 15:10 |
Miyagi | Soho (togatta spa) | Effectiveness of discriminative approaches for speech recognition under noisy environments on the 2nd CHiME Challenge Yuuki Tachioka (Mitsubishi Electric), Shinji Watanabe, Jonathan Le Roux, John R Hershey (MERL) SP2013-55 |
The 2nd CHiME challenge is a difficult two-microphone speech recognition task with non-stationary interference. We inves... [more] | SP2013-55 pp.13-18 |
| IBISML | 2012-03-12 11:25 |
Tokyo | The Institute of Statistical Mathematics | Fully Bayesian speaker clustering based on hierarchical structured Dirichlet process mixture model Naohiro Tawara, Tetsuji Ogawa (Waseda Univ.), Shinji Watanabe (NTT/MERL), Atsushi Nakamura (NTT), Tetsunori Kobayashi (Waseda Univ.) IBISML2011-90 |
We proposed a novel speaker clustering method by estimating the structure of a fully Bayesian utterance generative model... [more] | IBISML2011-90 pp.21-28 |
| SP, NLC, IPSJ-SLP [detail] | 2011-12-20 09:00 |
Tokyo | Simultaneous application of speaker adaptation and noise mixture model estimation for noise suppression Masakiyo Fujimoto, Shinji Watanabe, Tomohiro Nakatani (NTT) NLC2011-46 SP2011-91 |
In this paper, we propose a joint processing method for a model-based noise suppression that simultaneously achieves spe... [more] | NLC2011-46 SP2011-91 pp.113-118 |
|
| HPB (2nd) |
2011-09-10 | Hokkaido | Hakodate Community Design Center | Constructing Activity Models with User's Physical Characteristics Takuya Maekawa, Shinji Watanabe (NTT) |
This paper proposes a new activity recognition method that constructs activity models specific to an end user without ne... [more] | |
| SP | 2011-06-23 14:15 |
Aichi | Nagoya Univ. | [Invited Talk]
Recent topics in acoustic modeling for speech recognition
-- A machine learning perspective based on generative and discriminative approaches -- Shinji Watanabe (NTT) SP2011-31 |
Acoustic model research is one of the most active area in speech recognition, and there have been various progress. This... [more] | SP2011-31 pp.7-10 |
| EA, SIP, SP | 2011-05-12 10:50 |
Osaka | Ritsumeikan Univ. | A Robust On-line Estimation Method of Noise Mixture Model for Noise Suppression Masakiyo Fujimoto, Tomohiro Nakatani, Shinji Watanabe (NTT) EA2011-2 SIP2011-2 SP2011-2 |
In this paper, we propose a robust on-line estimation method of noise mixture model for the statistical model-based nois... [more] | EA2011-2 SIP2011-2 SP2011-2 pp.7-12 |
| NLC, SP (Joint) [detail] |
2010-12-20 16:30 |
Tokyo | National Olympics Memorial Youth Center | Noise suppression method based on noise bias-residual decomposition and optimization Masakiyo Fujimoto, Shinji Watanabe, Tomohiro Nakatani (NTT Corp.) NLC2010-18 SP2010-91 |
In this paper, we propose a non-stationary noise estimation method based on bias-residual component decomposition, and s... [more] | NLC2010-18 SP2010-91 pp.43-48 |
| SP, NLC | 2009-12-21 13:25 |
Tokyo | Univ. of Tokyo | Evaluation of Search Error Risk Minimization in Viterbi Beam Search Takaaki Hori, Shinji Watanabe, Atsushi Nakamura (NTT Corp.) NLC2009-15 SP2009-79 |
[more] | NLC2009-15 SP2009-79 pp.31-35 |
| PRMU | 2009-08-31 10:00 |
Miyagi | Tohoku Univ. | Geometric Margin for a General Class of Discriminant Functions and Its Control for Minimum Error Classification Hideyuki Watanabe (NICT), Shigeru Katagiri, Kouta Yamada (Doshisha Univ.), Erik McDermott, Atsushi Nakamura, Shinji Watanabe (NTT Corp.), Miho Ohsaki (Doshisha Univ.) PRMU2009-60 |
The recent dramatic growth of computation power has resulted in increased interests in discriminative training (DT) meth... [more] | PRMU2009-60 pp.1-6 |
| PRMU | 2009-09-01 09:00 |
Miyagi | Tohoku Univ. | Implementation and Experimental Evaluation of Ensemble Minimum Classification Error Training Shin'ichi Taniguchi (Doshisha University), Hideyuki Watanabe (NICT), Shigeru Katagiri, Kohta Yamada (Doshisha University), Atsushi Nakamura, Erik McDermott, Shinji Watanabe (NTT), Naho Nishijima, Miho Ohsaki (Doshisha Univ.) PRMU2009-67 |
Recently, we developed a noble Ensemble-based Minimum Classification Error training method (EMCE) by combining the advan... [more] | PRMU2009-67 pp.103-108 |
[Return to Top Page]