Masato Mimura
YOU?
Author Swipe
View article: Flux homomorphism and bilinear form constructed from Shelukhin's quasimorphism
Flux homomorphism and bilinear form constructed from Shelukhin's quasimorphism Open
Given a closed connected symplectic manifold $(M,ω)$, we construct an alternating $\mathbb{R}$-bilinear form $\mathfrak{b}=\mathfrak{b}_{μ_{\mathrm{Sh}}}$ on the real first cohomology of $M$ from Shelukhin's quasimorphism $μ_{\mathrm{Sh}}$…
View article: Non-extendablity of Shelukhin's quasimorphism and non-triviality of Reznikov's class
Non-extendablity of Shelukhin's quasimorphism and non-triviality of Reznikov's class Open
Shelukhin constructed a quasimorphism on the universal covering of the group of Hamiltonian diffeomorphisms for a general closed symplectic manifold. In the present paper, we prove the non-extendability of that quasimorphism for certain sy…
View article: Microphone Array Geometry Independent Multi-Talker Distant ASR: NTT System for the DASR Task of the CHiME-8 Challenge
Microphone Array Geometry Independent Multi-Talker Distant ASR: NTT System for the DASR Task of the CHiME-8 Challenge Open
In this paper, we introduce a multi-talker distant automatic speech recognition (DASR) system we designed for the DASR task 1 of the CHiME-8 challenge. Our system performs speaker counting, diarization, and ASR. It handles various recordin…
View article: Boosting Hybrid Autoregressive Transducer-based ASR with Internal Acoustic Model Training and Dual Blank Thresholding
Boosting Hybrid Autoregressive Transducer-based ASR with Internal Acoustic Model Training and Dual Blank Thresholding Open
A hybrid autoregressive transducer (HAT) is a variant of neural transducer that models blank and non-blank posterior distributions separately. In this paper, we propose a novel internal acoustic model (IAM) training strategy to enhance HAT…
View article: Alignment-Free Training for Transducer-based Multi-Talker ASR
Alignment-Free Training for Transducer-based Multi-Talker ASR Open
Extending the RNN Transducer (RNNT) to recognize multi-talker speech is essential for wider automatic speech recognition (ASR) applications. Multi-talker RNNT (MT-RNNT) aims to achieve recognition without relying on costly front-end source…
View article: NTT Multi-Speaker ASR System for the DASR Task of CHiME-8 Challenge
NTT Multi-Speaker ASR System for the DASR Task of CHiME-8 Challenge Open
We present a distant automatic speech recognition (DASR) system developed for the CHiME-8 DASR track. It consists of a diarization first pipeline. For diarization, we use end-to-end diarization with vector clustering (EEND-VC) followed by …
View article: NTT Multi-Speaker ASR System for the DASR Task of CHiME-8 Challenge
NTT Multi-Speaker ASR System for the DASR Task of CHiME-8 Challenge Open
International audience
View article: Sentence-wise Speech Summarization: Task, Datasets, and End-to-End Modeling with LM Knowledge Distillation
Sentence-wise Speech Summarization: Task, Datasets, and End-to-End Modeling with LM Knowledge Distillation Open
This paper introduces a novel approach called sentence-wise speech summarization (Sen-SSum), which generates text summaries from a spoken document in a sentence-by-sentence manner. Sen-SSum combines the real-time processing of automatic sp…
View article: SpeakerBeam-SS: Real-time Target Speaker Extraction with Lightweight Conv-TasNet and State Space Modeling
SpeakerBeam-SS: Real-time Target Speaker Extraction with Lightweight Conv-TasNet and State Space Modeling Open
Real-time target speaker extraction (TSE) is intended to extract the desired speaker's voice from the observed mixture of multiple speakers in a streaming manner. Implementing real-time TSE is challenging as the computational complexity mu…
View article: Invariant quasimorphisms and generalized mixed Bavard duality
Invariant quasimorphisms and generalized mixed Bavard duality Open
This article provides an expository account of the celebrated duality theorem of Bavard and three its strengthenings. The Bavard duality theorem connects scl (stable commutator length) and quasimorphisms on a group. Calegari extended the f…
View article: Coarse group theoretic study on stable mixed commutator length
Coarse group theoretic study on stable mixed commutator length Open
Let $G$ be a group and $N$ a normal subgroup of $G$. We study the large scale behavior, not the exact values themselves, of the stable mixed commutator length $scl_{G,N}$ on the mixed commutator subgroup $[G,N]$; when $N=G$, $scl_{G,N}$ eq…
View article: Time-domain Speech Enhancement Assisted by Multi-resolution Frequency Encoder and Decoder
Time-domain Speech Enhancement Assisted by Multi-resolution Frequency Encoder and Decoder Open
Time-domain speech enhancement (SE) has recently been intensively investigated. Among recent works, DEMUCS introduces multi-resolution STFT loss to enhance performance. However, some resolutions used for STFT contain non-stationary signals…
View article: End-to-End Generation of Written-style Transcript of Speech from Parliamentary Meetings
End-to-End Generation of Written-style Transcript of Speech from Parliamentary Meetings Open
従来の音声認識システムは，入力音声に現れるすべての単語を忠実に再現するように設計されているため，認識精度が高いときでも，人間にとって読みやすい文を出力するとは限らない．これに対して，本研究では，フィラーや言い誤りの削除，句読点や脱落した助詞の挿入，また口語的な表現の修正など，適宜必要な編集を行いながら，音声から直接可読性の高い書き言葉スタイルの文を出力する新しい音声認識のアプローチについて述べる．我々はこのアプローチを単一のニューラルネットワークを用いた音声から書き言葉への…
View article: Survey on invariant quasimorphisms and stable mixed commutator length
Survey on invariant quasimorphisms and stable mixed commutator length Open
A homogeneous quasimorphism $ϕ$ on a normal subgroup $N$ of $G$ is said to be $G$-invariant if $ϕ(gxg^{-1}) = ϕ(x)$ for every $g \in G$ and for every $x \in N$. Invariant quasimorphisms have naturally appeared in symplectic geometry and th…
View article: Non-autoregressive Error Correction for CTC-based ASR with Phone-conditioned Masked LM
Non-autoregressive Error Correction for CTC-based ASR with Phone-conditioned Masked LM Open
Connectionist temporal classification (CTC) -based models are attractive in automatic speech recognition (ASR) because of their non-autoregressive nature. To take advantage of text-only data, language model (LM) integration approaches such…
View article: Distilling the Knowledge of BERT for CTC-based ASR
Distilling the Knowledge of BERT for CTC-based ASR Open
Connectionist temporal classification (CTC) -based models are attractive because of their fast inference in automatic speech recognition (ASR). Language model (LM) integration approaches such as shallow fusion and rescoring can improve the…
View article: Invariant quasimorphisms for groups acting on the circle and non-equivalence of SCL
Invariant quasimorphisms for groups acting on the circle and non-equivalence of SCL Open
We construct invariant quasimorphisms for groups acting on the circle. Furthermore, we provide a criterion for the non-extendablity of the resulting quasimorphisms and an explicit formula which relates the values of our quasimorphisms to t…
View article: Mixed commutator lengths, wreath products and general ranks
Mixed commutator lengths, wreath products and general ranks Open
In the present paper, for a pair $(G,N)$ of a group $G$ and its normal subgroup $N$, we consider the mixed commutator length $\mathrm{cl}_{G,N}$ on the mixed commutator subgroup $[G,N]$. We focus on the setting of wreath products: $ (G,N)=…
View article: Synthesizing waveform sequence-to-sequence to augment training data for sequence-to-sequence speech recognition
Synthesizing waveform sequence-to-sequence to augment training data for sequence-to-sequence speech recognition Open
Sequence-to-sequence (seq2seq) automatic speech recognition (ASR) recently achieves state-of-the-art performance with fast decoding and a simple architecture. On the other hand, it requires a large amount of training data and cannot use te…
View article: ASR Rescoring and Confidence Estimation with ELECTRA
ASR Rescoring and Confidence Estimation with ELECTRA Open
In automatic speech recognition (ASR) rescoring, the hypothesis with the fewest errors should be selected from the n-best list using a language model (LM). However, LMs are usually trained to maximize the likelihood of correct word sequenc…
View article: The space of non-extendable quasimorphisms
The space of non-extendable quasimorphisms Open
For a pair $(G,N)$ of a group $G$ and its normal subgroup $N$, we consider the space of quasimorphisms and quasi-cocycles on $N$ non-extendable to $G$. To treat this space, we establish the five-term exact sequence of cohomology relative t…
View article: Commuting symplectomorphisms on a surface and the flux homomorphism
Commuting symplectomorphisms on a surface and the flux homomorphism Open
Let $(S,ω)$ be a closed connected oriented surface whose genus $l$ is at least two equipped with a symplectic form. Then we show the vanishing of the cup product of the fluxes of commuting symplectomorphisms. This result may be regarded as…
View article: Automatic Speech Recognition for the Archive of Ainu Folklores
Automatic Speech Recognition for the Archive of Ainu Folklores Open
本稿では，アイヌ民話（ウウェペケㇾ）の音声認識に関する我々の取り組みについて述べる．まず，2 つの博物館から提供されたアイヌ語アーカイブのデータを元に，沙流方言を対象としたアイヌ語音声コーパスを構築した．次に，このコーパスを用いて注意機構モデルに基づく音声認識システムを構成し，音素・音節・ワードピース・単語の 4 つの認識単位について検討した．その結果，音節単位での音声認識精度が最も高くなることがわかり，話者クローズド条件と話者オープン条件のそれぞれについて，音素認識精度で…
View article: Constellations in prime elements of number fields
Constellations in prime elements of number fields Open
Given any number field, we prove that there exist arbitrarily shaped constellations consisting of pairwise non-associate prime elements of the ring of integers. This result extends the celebrated Green-Tao theorem on arithmetic progression…
View article: CTC-Synchronous Training for Monotonic Attention Model
CTC-Synchronous Training for Monotonic Attention Model Open
Monotonic chunkwise attention (MoChA) has been studied for the online streaming automatic speech recognition (ASR) based on a sequence-to-sequence framework. In contrast to connectionist temporal classification (CTC), backward probabilitie…
View article: Enhancing Monotonic Multihead Attention for Streaming ASR
Enhancing Monotonic Multihead Attention for Streaming ASR Open
We investigate a monotonic multihead attention (MMA) by extending hard monotonic attention to Transformer-based automatic speech recognition (ASR) for online streaming applications. For streaming inference, all monotonic attention (MA) hea…
View article: End-to-end Music-mixed Speech Recognition
End-to-end Music-mixed Speech Recognition Open
Automatic speech recognition (ASR) in multimedia content is one of the promising applications, but speech data in this kind of content are frequently mixed with background music, which is harmful for the performance of ASR. In this study, …
View article: Distilling the Knowledge of BERT for Sequence-to-Sequence ASR
Distilling the Knowledge of BERT for Sequence-to-Sequence ASR Open
Attention-based sequence-to-sequence (seq2seq) models have achieved promising results in automatic speech recognition (ASR). However, as these models decode in a left-to-right way, they do not have access to context on the right. We levera…