Meta Reality Labs Research is looking for upcoming scientists and researchers with great interest in designing, modeling, and prototyping advanced vision systems for real-time, low-power processing. The goal of this role is to optimize the full-stack vision system spanning from the image sensor, device processors, cloud computation, and corresponding algorithms, to help people capture the moments that matter with smart glasses. Ideal candidates will have a comprehensive understanding of computational imaging, distributed computation, computer vision, large language models, and machine learning algorithms.
Our internships are twelve (12), to twenty-four (24) weeks long and we have various start dates throughout the year.
Responsibilities
- Research on design / model / execution of vision and multimodal ML algorithms for low-power smart-glasses camera systems.
- Develop methods that connect large language models with on-device and in-sensor processing.
- Build datasets, benchmarks, and evaluation protocols for egocentric capture tasks.
- Prototype and demonstrate novel capture experiences on smart-glasses camera hardware.
- Collaboration with and support of other researchers and engineers across various disciplines.
- Communication of research agenda, progress and results.
Minimum qualifications
- Currently has, or is in the process of obtaining, a PhD in the fields of Computer Science, Electrical Engineering, Machine Learning, or related field
- 2+ years of coding experience with programming languages such as Python/C++
- Advanced experience with computer vision (CV) and modern deep learning research, including algorithm development and optimization
- Experience with video understanding or vision-language models
- Must obtain work authorization in the country of employment at the time of hire and maintain ongoing work authorization during employment
Preferred qualifications
- Experience with lightweight machine learning on embedded systems, including model optimization and deployment
- Experience with LLMs and their APIs is a strong plus
- Experience with egocentric video or the full machine vision pipeline spanning cameras, ISPs, and high-level algorithms
- Proven track record of achieving significant results in CV and ML research, as demonstrated by publications, grants, or fellowships
- Intent to return to degree-program after the completion of the internship/co-op
Job description provided by Meta.
View the original posting