Senior Perception Engineer – Robotics Vision & SLAM Lead

XDOF; San Mateo; On-site; PHP 1,200,000 - 2,400,000; Full time

Job summary

XDOF in the Philippines is seeking an experienced robotics perception engineer to advance our perception stack. You will work on hand pose estimation, 3D vision, SLAM localization, and real-time deployment on embedded platforms, directly shaping the quality of our robot training data.

You will design and optimize perception models, build auto‑annotation pipelines, and collaborate with ML engineers and hardware teams to deliver robust, production‑grade systems.

Qualifications

• 5+ years of industry experience in robot perception or computer vision.
• Strong 3D vision fundamentals and SLAM experience.
• Hands-on pose estimation with hand/body models.
• Proficiency in DL frameworks for perception model training and evaluation.
• CUDA/TensorRT deployment on embedded platforms (Jetson, Horizon, etc.).
• Proficient in C++ and Python with ROS/ROS2 development.
• Experience with AI coding agents is a plus.

Responsibilities

• Design and optimize hand pose and full-body pose estimation pipelines.
• Develop and maintain multi-camera calibration, SLAM, and visual localization systems.
• Train and deploy perception models for object detection, segmentation, and 6DoF pose estimation.
• Write custom CUDA kernels for acceleration and real-time inference.
• Build automated annotation pipelines and QA modules for data quality assurance.
• Collaborate with ML and data infrastructure teams to meet downstream requirements.
• Contribute to evaluation frameworks tracking model performance vs data quality.

Skills

5+ years experience

3D vision

SLAM frameworks

pose estimation

DL training frameworks

TensorRT/CUDA

CUDA kernels

C++ & Python

ROS/ROS2

AI coding agents

Tools

ROS/ROS2

TensorRT

CUDA

Job description

XDOF in the Philippines is seeking an experienced robotics perception engineer to advance our perception stack. You will work on hand pose estimation, 3D vision, SLAM localization, and real-time deployment on embedded platforms, directly shaping the quality of our robot training data.

You will design and optimize perception models, build auto‑annotation pipelines, and collaborate with ML engineers and hardware teams to deliver robust, production‑grade systems.