X

XDOF

Data pipelines and annotation for robotics

Member of Technical Staff - Perception

Computer VisionConfirmed live in the last 24 hours
Full-timeSan Mateo, CA, USA$180k - $300k /yrHybrid

About the Job

Requirements

  • At least 5 years of industry experience in robot perception or computer vision.
  • Strong 3D vision fundamentals, including stereo and structured-light camera principles and 3D reconstruction.
  • Proficiency with SLAM frameworks such as ORB-SLAM, VINS-Mono, or FastLIO, or experience developing V-SLAM systems.
  • Hands-on engineering experience with human pose estimation, including hand joints using MediaPipe or MANO, or full-body pose using OpenPose or SMPLify.
  • Proficiency with deep learning training frameworks for perception model training, tuning, and evaluation.
  • Experience deploying TensorRT and optimizing real-time inference on embedded platforms such as Jetson or Horizon.
  • CUDA programming fundamentals and ability to write or debug custom kernels.
  • Proficiency in C++ and Python, with ROS or ROS2 development experience.
  • Proficiency with AI coding agents.

Responsibilities

  • Design and optimize hand pose estimation pipelines to extract accurate joint angles from teleoperation data collection.
  • Build full-body pose estimation systems for motion capture and teleoperation action annotation ground-truth generation.
  • Research and apply markerless, vision-based pose estimation methods to reduce data collection costs.
  • Retarget pose estimation outputs with robot joint-angle data to generate consistent training annotations.
  • Design and maintain intrinsic and extrinsic calibration pipelines for multi-camera arrays, including factory calibration and online recalibration.
  • Build visual SLAM or V-SLAM systems for real-time localization and scene reconstruction on data collection platforms.
  • Implement hand-eye calibration between cameras and robot end-effectors.
  • Develop temporal alignment solutions across multimodal sensors, including cameras, IMUs, data gloves, and force sensors.
  • Train and iterate on perception models, including object detection, instance segmentation, and 6DoF pose estimation.
  • Optimize model inference using TensorRT and CUDA for real-time performance on robot embedded platforms.
  • Write custom CUDA kernels for low-level acceleration of perception tasks.
  • Design evaluation metric frameworks for perception models and track the relationship between model performance and data quality.
  • Contribute to designing automated annotation pipelines that convert sensor data into structured training labels.
  • Build automated quality-assurance modules to filter low-quality data, including anomalous frames, failed demonstrations, and sensor dropouts.
  • Collaborate with machine learning engineers and data infrastructure teams to ensure perception output formats meet downstream vision-language-action model training requirements.
  • Establish feedback mechanisms connecting perception accuracy to model training outcomes and continuously improve annotation quality.

Desired Qualifications

  • Engineering experience with 6DoF object pose estimation, using methods such as FoundPose, FoundationPose, or GDR-Net.
  • Familiarity with 3D Gaussian Splatting or NeRF for scene reconstruction or data augmentation.
  • Experience with robot manipulation or teleoperation systems.
  • End-to-end development experience with automated annotation pipelines or ground-truth generation systems.
  • Published research in perception, pose estimation, or robotics.

About the Company

XDOF

Data pipelines and annotation for robotics

XDOF builds data pipelines, data-collection tools, and annotation systems that frontier AI labs and robotics companies use to train robots to operate in the physical world. It tackles the data bottleneck in physical AI by combining teleoperation, simulation, and egocentric data collection, then layering in data cleaning, tooling, and annotation to prepare usable datasets. The company differentiates itself through its end-to-end data infrastructure for physical AI, its mix of teleoperation, simulation, and first-person data methods, and its Berkeley-founded background from 2024. Its goal is to enable robots to learn to operate in the real world by providing the pipelines and labels needed for training.

Company size
1-10
Company stage
Series A
Total funding
$78.3M
Headquarters
Berkeley, CA
Founded
2024

Recently Posted Jobs

Explore more opportunities at XDOF

See more jobs