Now, CAPRICEP made an objective comparison of pitch extractors possible. The right bottom extractor shows outstanding performance. It is the re-engineered version of my Google UK extractor. I will put it on GitHub.
https://t.co/iiHuRxyUPf
https://t.co/LSO2yBEKk3 https://t.co/CcHNQiFMSu
K. Matsubara et al., "Comparison of real-time multi-speaker neural vocoders on CPUs" now officially published. Multi-speaker HiFi-GAN with LPCNet features is investigated and compared with MWDLP.
https://t.co/ROSUkLapBw
PatrickくんのD論 (2020.3) / High-Quality and Flexible Voice Conversion Techniques based on Statistical Spectral and Waveform Modeling https://t.co/bQgXiuaUw7
Happy New Year!! K. Matsubara's work in NICT "Investigation of training data size for real-time neural vocoders on CPUs," in AST is now online as open access!!
https://t.co/9k3XWpTY96
https://t.co/dBxQ5lPtiH
Joint Adversarial Training of Speech Recognition and Synthesis Models for Many-to-One Voice Conversion Using Phonetic Posteriorgrams https://t.co/VUAyVh2w2S
論文が公開されました。 > K. Yamamoto et al., “Speech intelligibility prediction with the dynamic compressive gammachirp filterbank and modulation power spectrum,” Acoustical Science and Technology, Vol. 40, Issue 2, pp. 84-92, 2019. https://t.co/WecgS3FAN8
FYI / Okamoto et al., "Deep neural network-based power spectrum reconstruction to improve quality of vocoded speech with limited acoustic parameters" https://t.co/Dgh2lZhz3O
「Constructing text-to-speech systems for languages with unknown pronunciations」と題する論文がAcoustical Science and Technologyに掲載されましたhttps://t.co/uHoUkLgSxf。
音素等の発音が未知の言語に対して音声合成システムを構築する手法について論じています。