Data Annotator (LiDAR)

Sequential Technologies International; Legazpi; Presencial; PHP 350.000 - 600.000; Jornada completa

Descripción de la vacante

Sequential Technologies International is seeking a meticulous LiDAR Data Annotator to join our AI Operations team in Legazpi. You will analyze, label, and segment complex 3D point cloud data from LiDAR sensors to train autonomous systems and spatial AI applications.

The role emphasizes 3D bounding box annotations, semantic segmentation, and cross-checking with camera data, requiring strong spatial awareness and centimeter-level precision.

Formación

• 1–3+ years of hands-on experience annotating 3D LiDAR point clouds or multi-modal sensor datasets.
• Experience with industry-standard 3D annotation platforms is preferred.
• Strong spatial reasoning and attention to centimeter-level accuracy.

Responsabilidades

• 3D Point Cloud Labeling: perform 3D bounding box annotations (cuboids), semantic segmentation, and polyline tracking on multi-sensor fused dataset (LiDAR, Radar, Camera).
• Sensor Fusion Alignment: cross-reference 3D LiDAR point clouds with 2D camera footage to ensure accurate depth perception, object detection, and tracking consistency.
• Quality & Accuracy: consistently meet or exceed strict accuracy benchmarks across complex edge cases.
• Tooling Feedback: test internal or third-party annotation platforms and report bugs, edge cases, or UX enhancements to machine learning engineers.
• Edge Case Analysis: identify, flag, and document edge cases for review by QA leads and ML researchers.
• Collaboration: participate in calibration sessions to align on ambiguous edge cases and updated labeling guidelines.

Conocimientos

LiDAR data annotation

3D spatial reasoning

Attention to detail

Herramientas

CVAT

Scale AI

Labelbox

Kili Technology

Appen

Deepen AI

Playment

CloudCompare

MeshLab

ROS (rviz)

3D Bounding Box Editors

Descripción del empleo

Job Title: Data Annotator – LiDAR & 3D Point Cloud

Job Summary

We are seeking a detail-oriented and tech-savvy LiDAR Data Annotator to join our AI Operations team. In this role, you will analyze, label, and segment complex 3D point cloud data generated by LiDAR sensors. Your primary focus will be annotating 3D bounding boxes and segmenting objects (e.g., pedestrians, vehicles, road infrastructure, vegetation) to train computer vision models for autonomous systems and spatial AI applications.

Key Responsibilities

• 3D Point Cloud Labeling: Perform 3D bounding box annotations (cuboids), semantic segmentation, and polyline tracking on multi-sensor fused dataset (LiDAR, Radar, Camera).

• Sensor Fusion Alignment: Cross-reference 3D LiDAR point clouds with 2D camera footage to ensure accurate depth perception, object detection, and tracking consistency.

• Quality & Accuracy: Consistently meet or exceed strict accuracy benchmarks (e.g., box fitting, orientation, rotation, occlusions) across complex edge cases.

• Tooling Feedback: Test internal or third-party annotation platforms and report bugs, edge cases, or UX enhancements to machine learning engineers.

• Edge Case Analysis: Identify, flag, and document edge cases (e.g., sparse point clouds, sensor noise, weather reflections) for review by QA leads and ML researchers.

• Collaboration: Participate in calibration sessions to align on ambiguous edge cases and updated labeling guidelines.

Required Qualifications & Background

• Experience: 1–3+ years of hands-on experience annotating 3D LiDAR point clouds or multi-modal sensor datasets.

• Software Proficiency: Strong proficiency with industry-standard 3D annotation platforms such as:

  • Commercial / Cloud Platform: CVAT, Scale AI, Labelbox, Kili Technology, Appen, Deepen AI, Playment.

  • Open Source / Local: CloudCompare, MeshLab, ROS (rviz), 3D Bounding Box Editors.

• Spatial Reasoning: Excellent 3D spatial awareness, object orientation understanding, and depth perception capabilities.

• Hardware Understanding: Familiarity with how LiDAR sensors work (e.g., Velodyne, Ouster, Hesai, Luminar) and common point cloud artifacts (intensity, density variations, occlusion).

• Attention to Detail: Ability to spot minute patterns and correct precise spatial measurements down to centimeter-level accuracy.