Senior Perception Engineer: Robotics Vision Architect

XDOF; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Benefits offered by this job

High autonomy

Elite robotics team

Proprietary hardware platforms

Fast-paced environment

Job summary

XDOF seeks a senior robotics perception engineer in San Francisco to lead end‑to‑end perception development. You will design and optimize pose estimation, calibrate multi‑camera rigs, and deploy real‑time perception models on embedded platforms.

Strong background in 3D vision, SLAM, and CUDA is required, with ROS/ROS2 fluency and collaboration with ML teams. You will drive data collection loops, annotation pipelines, and model evaluation, shaping the quality of training data for advanced

Qualifications

• 5+ years in robot perception or computer vision with hands‑on delivery.
• Strong 3D vision fundamentals and 3D reconstruction knowledge.
• Experience with SLAM frameworks or V‑SLAM system development.
• Hands‑on pose estimation with markerless methods or full‑body tech.
• Proficient in DL training for perception tasks and model evaluation.
• CUDA programming basics; ability to write/debug kernels.
• Proficient in C++ and Python with ROS/ROS2 development.
• Familiar with AI tools and agent coding workflows.

Responsibilities

• Design and optimize hand and full‑body pose estimation pipelines.
• Build 3D perception and calibration for multi‑camera systems.
• Develop real‑time perception models and deploy with TensorRT/CUDA.
• Create end‑to‑end annotation pipelines from sensor data to labels.
• Collaborate with ML and data infra teams to align formats.

Skills

Robot perception

3D vision fundamentals

SLAM frameworks

Pose estimation

Deep learning for perception

CUDA

C++ & Python

ROS/ROS2

AI coding agents

Tools

TensorRT

OpenPose

MediaPipe

SMPLify

ORB-SLAM/V-SLAM

VINS-Mono

Job description

XDOF seeks a senior robotics perception engineer in San Francisco to lead end‑to‑end perception development. You will design and optimize pose estimation, calibrate multi‑camera rigs, and deploy real‑time perception models on embedded platforms.

Strong background in 3D vision, SLAM, and CUDA is required, with ROS/ROS2 fluency and collaboration with ML teams. You will drive data collection loops, annotation pipelines, and model evaluation, shaping the quality of training data for advanced