Hiroki Kanagawa (research engineer @ NTT Corp.)

Affiliation

Hiroki Kanagawa (金川裕紀) is research engineer at NTT Corporation. The detail CV is in preparation.

Work experience

Dec, 2017 - Now: Research Engineer, Human Insight Laboratory, NTT Human Informatics Laboratories, NTT Corporation, Japan.
Apr, 2013 - Nov, 2017: Research engineer at Information Technology R&D Center, Mitsubishi Electric Corporation, Japan.

Education

M.E. in information processing from Tokyo Institute of Technology in 2013, Japan (Supervisor: Takao Kobayashi)
B.E. in electronics engineering from The University of Electro-Communications in 2011, Japan (Supervisor: Tetsuo Kirimoto)

Publication

Publication is here

Research interest

text-to-speech synthesis, voice conversion, neural vocoder, machine learning, automatic speech recognition

Programming Skill

C, C++ (good at fast implementing using e.g. SIMD), C++/CLI, C#, Perl, Python (over 10 years), WPF, ASP.net, Javascript

Hobby

Bonsai 🪴, Playing/watching baseball ⚾, Watching anime 😎

Publication

Anonymous published on 2023-10-04

Peer-reviewed paper 2024 Hiroki Kanagawa, Takafumi Moriya, Yusuke Ijima, “Pre-training Neural Transducer-based Streaming Voice Conversion for Faster Convergence and Alignment-free Training,” Proc. Interspeech 2024, pp. 2755-2759, Aug. 2024 @ Kos, Greece. Hiroki Kanagawa and Yusuke Ijima, “Knowledge Distillation from Self-Supervised Representation Learning Model with Discrete Speech Units for Any-to-Any Streaming Voice Conversion,” Proc. Interspeech 2024, pp. 4393-4397, Aug. 2024 @ Kos, Greece. 2023 Hiroki Kanagawa, Takafumi Moriya, Yusuke Ijima, “VC-T: Streaming Voice Conversion Based on Neural Transducer,” Proc.