Real-time Voice Conversion AI chip

AI chip that can synthesize and convert speech in real time using FPGA

Voice Conversion Engine
Speech Synthesis Engine
Real-time Voice Conversion AI chip


We are researching and developing an AI chip that performs high-quality speech synthesis in real time using FPGAs. This will enable the practical application of speech synthesis technology for applications that require real-time performance, such as human speech, which has been difficult to achieve with conventional technologies due to the delay involved.

Use Cases

For example, by implementing an AI that has learned the voice of a certain character on our chip, anyone will be able to speak in that character's voice in real time. This is expected to create new, unprecedented live entertainment content and provide voice content that does not depend on human resources.

101 Nefrock Lab Ookayama, 3-35-5 Kitasenzoku, Ootaku-ku, Tokyo 145-0062 Japan