Audio Quality Research Intern — Speech & AI

Meta; Cambridge; On-site; GBP 21,000 - 30,000; Full time

Job summary

Meta Reality Labs Research invites you to join our audio team as an intern focused on speech perception and audio quality. The project targets degradation causes, phonemic error patterns, and the role of multimodal LLMs in evaluating quality.

A twelve to twenty-four week program offers multiple start dates throughout the year. You will work with PhD candidates in Speech and Hearing Science, Auditory Neuroscience, Computational Neuroscience, CS, AI, Generative AI, Transformer models, and ML,

Qualifications

• Candidate currently pursuing a PhD in Speech & Hearing Science, Auditory Neuroscience, Computational Neuroscience, CS, AI, or ML.
• Proficiency with Python, MATLAB, or similar languages.
• Experience with ML platforms such as PyTorch or TensorFlow.
• Background in speech perception, psychoacoustics, or acoustic phonetics.
• Experience deploying novel audio models and LLMs.
• Must have work authorization in the country of employment.
• Experience building audio ML models and LLMs.
• Demonstrated software engineering activity (internships, projects, open source).
• Experience advancing AI methods in CV or audio processing.
• Experience with audio/speech quality assessment and multichannel processing.
• Strong data analysis from large, complex datasets.
• Publication record at top venues is a plus.
• Ability to work cross-functionally and collaborate in teams.
• Intent to return to a degree program after the internship.

Responsibilities

• Investigate why processed speech sounds degraded or robotic.
• Identify systematic phonemic errors affecting perceived quality.
• Explore audio vs video LLM capabilities for speech tasks.
• Collaborate with cross-functional teams on research questions.
• Present findings through reports and demonstrations.

Skills

PhD candidate

Python

ML frameworks

Speech perception

LLMs

Open source

Team collaboration

Communication

Data analysis

Research experience

Education

PhD in field

Tools

PyTorch

TensorFlow

MATLAB

GitHub

Job description

Meta Reality Labs Research invites you to join our audio team as an intern focused on speech perception and audio quality. The project targets degradation causes, phonemic error patterns, and the role of multimodal LLMs in evaluating quality.

A twelve to twenty-four week program offers multiple start dates throughout the year. You will work with PhD candidates in Speech and Hearing Science, Auditory Neuroscience, Computational Neuroscience, CS, AI, Generative AI, Transformer models, and ML,