Skip to content

R&D Internship 2027 — six audio projects (source separation, DSP, acoustics, music AI)

Yamaha Hamamatsu / Yokohama, Japan

Apply at source
Salary
—
Type
—
Level
internship
Posted
2 months ago

Yamaha Corporation is recruiting R&D engineers worldwide for a paid internship in Japan, based at Yamaha's Innovation Center. Term: approximately three months, Summer to Autumn 2027. Location: Hamamatsu / Yokohama, Japan. Requirement: doctoral or master's students in engineering, computer science, mathematics or related fields (master's or senior students for the Product Development project). Conditions: paid, with a stipend, visa application support, transportation cost assistance and furnished accommodation. Applications open 28 September 2026; due 18 November 2026 (AOE, UTC-12). Selection is an application followed by an online interview. Six projects are offered. Five sit in the Research & Development Division and one in the Product Development Division: 1. Conditional Accompaniment Generation for Free-Tempo Musical Performance (Research & Development Division) Recent advances in deep learning have enabled high-quality music generation. However, generating accompaniment that follows a musician’s expressive, free-tempo performance in real time remains a challenging problem. In this internship, you will develop a conditional accompaniment generation system guided by user performance. By using musical scores and/or reference audio as conditions, the system will generate accompaniment for a specific musical piece while adapting to the user’s timing and expression. The project will involve investigating low-latency music generation and performance-following techniques for human-AI ensemble performance. Required: Experience with deep learning frameworks 2. Development of Ultra-Low-Latency Music Source Separation (Research & Development Division) This project aims to develop core technologies for next-generation music production and musical experiences through ultra-low-latency music source separation. You will explore machine learning and audio signal processing techniques to achieve high-quality source separation with minimal processing delay. While the project addresses music source separation in general, a primary focus will be on extracting clean vocal signals by suppressing bleed, acoustic feedback, and other unwanted components. You will develop and evaluate source separation models, as well as analyze the trade-offs among latency, computational efficiency, and separation performance. In addition, based on these investigations 3. Modeling Piano Key Force Inputs and Their Application to Inverse Dynamics Analysis of Piano Performance (Research & Development Division) Pianists generate a wide range of sounds through subtle variations in the force applied to the keyboard. Modeling these control inputs can contribute to improved instrument design, performance analysis, and feedback systems for piano education. This project aims to estimate and model force inputs to piano keys using multimodal instrument and performer sensing, including force transducers, key and hammer motion sensing, and motion capture data. You will develop machine learning models that infer force-related control variables from sensor measurements and, where feasible, integrate these estimates into inverse dynamics analyses to better understand the biomechanics of piano performance. Requi 4. Gray-box Modeling for Audio Equipment and Musical Instruments (Research & Development Division) Audio equipment and musical instruments are complex systems in which mechanical vibration, acoustics, electric circuits, and performance inputs interact, often with significant nonlinear behavior. This internship focuses on gray-box modeling, which combines physical and mathematical prior knowledge with data-driven modeling techniques, such as machine learning-based system identification, parameter estimation, and inverse modeling. Target systems may include loudspeakers, amplifiers, wind instruments, and FM synthesizers. The project supports behavior prediction, control, and model-based design through accurate and interpretable modeling. Using experimental facilities and measured data from 5. Enhancing Subjective Evaluation Models for Piano Sound Using Performance Data (Research & Development Division) We have developed subjective evaluation models for piano sound by collecting and systematically organizing the words pianists use to describe sound in interviews and performance evaluations. While these language-based models provide a solid foundation for capturing how pianists perceive and evaluate sound, certain aspects are difficult to fully represent through words alone. We therefore aim to further refine these models by incorporating additional performance data, such as MIDI recordings and video footage. In this internship, you will use a multimodal measurement system to record piano performance data, including MIDI, video, and audio. You will then extract relevant features from the col 6. Development of Digital Effects for Synthesizers & Keyboards (Product Development Division) Even today, simulating analog effects in digital musical instruments and developing high-performance effect algorithms in the digital domain remain essential. During this internship, you will participate in development work that includes the following tasks, contributing to the improvement of the quality of future Yamaha products. Prototyping VST3 Digital Effects for Synthesizers and Keyboards: - understanding prototyping framework and development setup - researching and designing digital audio effects algorithm - realtime audio signal processing implementation as VST3 plugin - sound quality verification and value proposition validation Required: Experience in designing DSP algorithm Experie Note: Yamaha publishes these projects on a single page with no per-project application links, so they are listed here as one entry. Open the URL to see the full requirements for each project.

Scraped Sep 3, 2026 · Community submission, reviewed by moderators · expires Dec 2, 2026

R&D Internship 2027 — six audio projects (source separation, DSP, acoustics, music AI)

What's going on?

R&D Internship 2027 — six audio projects (source separation, DSP, acoustics, music AI)

Ask the team to put this role on the front page. Pick how long, and if you're the one hiring, add a note candidates should see — the team reviews every request and may edit it.

Help us improve ASoundJob

What's going on?