Zobrazeno 1 - 10
of 265
pro vyhledávání: '"Tetsuya TAKIGUCHI"'
Publikováno v:
IEEE Access, Vol 12, Pp 36990-36999 (2024)
In this paper, we investigate the use of the spontaneous speech of dysarthric people for training an automatic speech recognition (ASR) model for them. Although the spontaneous speech of dysarthric people can be collected relatively easily compared t
Externí odkaz:
https://doaj.org/article/51e225758a6e43888c296ceb14e6dc71
Autor:
Haruki Yamashita, Takuma Okamoto, Ryoichi Takashima, Yamato Ohtani, Tetsuya Takiguchi, Tomoki Toda, Hisashi Kawai
Publikováno v:
IEEE Access, Vol 12, Pp 31409-31421 (2024)
Although end-to-end (E2E) text-to-speech (TTS) models with HiFi-GAN-based neural vocoder (e.g. VITS and JETS) can achieve human-like speech quality with fast inference speed, these models still have room to further improve the inference speed with a
Externí odkaz:
https://doaj.org/article/65f4b9e442c0485f987d349b630c164a
Autor:
Koichi Tsunoda, Toyota Ishii, Hiroyuki Kuroda, Hiroaki Nakatani, Masaru Tateda, Sawako Masuda, Tetsuya Takiguchi, Fujinobu Tanaka, Hayato Misawa, Masamitsu Senarita, Mihiro Takazawa, Kenji Itoh, Thomas Baer
Publikováno v:
Heliyon, Vol 10, Iss 4, Pp e25751- (2024)
We speculated that increased blood-plasma levels of Substance P may serve as an indicator of glottal incompetence, which is usually indicated by reduced maximum phonation time. We performed an initial study to test the plausibility of this hypothesis
Externí odkaz:
https://doaj.org/article/ecddf9d57fb84fec8093bcebfc502327
Publikováno v:
EURASIP Journal on Image and Video Processing, Vol 2023, Iss 1, Pp 1-30 (2023)
Abstract Training Convolutional Neural Networks (CNN) is a resource-intensive task that requires specialized hardware for efficient computation. One of the most limiting bottlenecks of CNN training is the memory cost associated with storing the activ
Externí odkaz:
https://doaj.org/article/c22705eb6df54be494103044d0fa92b6
Publikováno v:
Remote Sensing, Vol 16, Iss 1, p 170 (2023)
Evapotranspiration (E) is one of the most uncertain components of the global water cycle (WC). Improving global E estimates is necessary to improve our understanding of climate and its impact on available surface water resources. This work presents a
Externí odkaz:
https://doaj.org/article/5a52db3f98d94eb9873c30b8aa47a3e7
Autor:
Hideki Mutai, Yukihide Momozawa, Yoichiro Kamatani, Atsuko Nakano, Hirokazu Sakamoto, Tetsuya Takiguchi, Kiyomitsu Nara, Michiaki Kubo, Tatsuo Matsunaga
Publikováno v:
Orphanet Journal of Rare Diseases, Vol 17, Iss 1, Pp 1-17 (2022)
Abstract Background Heterogeneous genetic loci contribute to hereditary hearing loss; more than 100 deafness genes have been identified, and the number is increasing. To detect pathogenic variants in multiple deafness genes, in addition to novel cand
Externí odkaz:
https://doaj.org/article/ea37ea9188d24a9bb98645768f046570
Publikováno v:
APSIPA Transactions on Signal and Information Processing, Vol 12, Iss 1 (2023)
Externí odkaz:
https://doaj.org/article/263c0415235144f4a2ac61bc48b1c317
Autor:
Yuki Takashima, Ryoichi Takashima, Ryota Tsunoda, Ryo Aihara, Tetsuya Takiguchi, Yasuo Ariki, Nobuaki Motoyama
Publikováno v:
EURASIP Journal on Audio, Speech, and Music Processing, Vol 2021, Iss 1, Pp 1-9 (2021)
Abstract We present an unsupervised domain adaptation (UDA) method for a lip-reading model that is an image-based speech recognition model. Most of conventional UDA methods cannot be applied when the adaptation data consists of an unknown class, such
Externí odkaz:
https://doaj.org/article/22667a463fb5441488e5b523b7cff10b
Autor:
Keisuke Matsubara, Takuma Okamoto, Ryoichi Takashima, Tetsuya Takiguchi, Tomoki Toda, Yoshinori Shiga, Hisashi Kawai
Publikováno v:
IEEE Access, Vol 9, Pp 94923-94933 (2021)
This paper investigates a real-time neural speech synthesis system on CPUs that can synthesize high-fidelity 48 kHz speech waveforms to cover the entire frequency range audible by human beings. Although most previous studies on 48 kHz speech synthesi
Externí odkaz:
https://doaj.org/article/2b1161e52b18489e9d7ce9bb4a6bde4d
Publikováno v:
EURASIP Journal on Audio, Speech, and Music Processing, Vol 2019, Iss 1, Pp 1-11 (2019)
Abstract Voice conversion (VC) is a technique of exclusively converting speaker-specific information in the source speech while preserving the associated phonemic information. Non-negative matrix factorization (NMF)-based VC has been widely researche
Externí odkaz:
https://doaj.org/article/39974abf0fd64b529c5abce02697237f