Arm
Staff Perception Scientist
$209K–$283K
Seattle, Washington · Staff · Full-time · Posted Sep 9, 2026
- Experience
- 6+ years in a relevant field; 3+ years moving perception research into real-time or production systems.
- Education
- PhD or equivalent experience is preferred; no required education stated.
- Visa sponsorship
- Relocation package including visa sponsorship assistance is offered for candidates who need it.
About the job
Researches and develops robust perception methods for Physical AI, then optimizes and deploys them on Arm-powered devices for real-time robotic and autonomous systems.
What you'll do
- Develop and evaluate perception algorithms for detection, segmentation, tracking, localization, scene understanding, and sensor fusion.
- Define datasets, metrics, and reproducible experiments; analyze failures and improve models.
- Optimize and integrate perception systems for on-device performance, while providing technical leadership and mentoring.
What they ask for
- 6+ years in applied perception, computer vision, machine learning, or a related field.
- Proficiency in Python and C++ with PyTorch, JAX, TensorFlow, or similar frameworks.
- Strong understanding of deep learning for perception, probability, optimization, 3D geometry, sensor fusion, efficient inference, and edge deployment.
- 3+ years moving perception research into real-time or production systems, plus a record of impact through deployed systems, publications, patents, or widely used research artifacts.
Nice to have
- PhD or equivalent experience in a related field.
- Experience in areas such as 3D perception, SLAM, multimodal learning, synthetic data, model compression, accelerators, or robotics.
Pay
$209,100-$282,900 per year (USD, base salary)
Tools
Python, C++, PyTorch, JAX, TensorFlow, Arm CPUs
AI focus
Develops perception AI for physical-world systems, including multimodal sensing, robotics, and autonomous decision-making, with emphasis on efficient on-device deployment.
Model-written summary · the original posting is authoritative