3D与LiDAR数据标注分析师
3D & LiDAR Data Annotation Analyst
职位描述
我们正在构建高质量的真实数据集，用于训练和验证在复杂现实环境中工作的自动驾驶和人工辅助机器的感知模型。
作为数据标注分析师，您将处理摄像头图像、视频和LiDAR点云，创建精确的标签，帮助感知系统识别物体、人员、地形和周围环境。您将使用专门的标注工具，遵循详细规范，同时平衡准确性、一致性和生产量。
您的影响
- 在图像、视频和LiDAR点云中对2D和3D物体进行标注，包括车辆、机械、人员、标志和环境特征。
- 在序列中创建并维护准确的分割、物体跟踪、姿态和关键点标注。
- 应用详细的标注指南，包括类别定义、遮挡规则、物体阈值以及包含/排除标准。
- 审核您的工作，响应质量保证反馈，并保持既定的质量和准确度标准。
- 识别并上报模糊场景、传感器伪影、工具问题和规范缺失，而不是做出假设。
- 清晰详细地记录数据质量和工具问题，并提供重现步骤。
- 参与校准会议、指南审查和团队站会，以保持一致的解释。
- 分享有助于改进标注指南、分类体系和边缘案例文档的观察结果。
在此职位上取得成功意味着在预期的产量下产出准确且一致的标注，同时帮助维护可靠且高质量的数据集。
您带来的能力
- 对细节有很强的关注力，能够在长时间内执行重复性、视觉密集型工作时保持准确性。
- 强大的空间推理能力，包括能够解读3D环境并理解物体的大小、位置和方向。
- 能够学习和操作复杂的专用软件，并熟练使用键盘快捷键和其他工具功能。
- 有在快节奏、可扩展环境中工作的经验，有明确的生产力、质量或准确度目标。
- 有图像标注生成式AI（GenAI）工作流程的经验。
- 可靠的高速互联网连接、无干扰的工作环境，以及能够在受控环境中使用公司提供的设备工作。
加分项
- 专科或学士学位。
- 相关经验
查看英文原文
About the Role
We’re building high-quality ground-truth datasets used to train and validate perception models for autonomous and operator-assisted machines working in complex, real-world environments.
As a Data Annotation Analyst, you’ll work with camera imagery, video, and LiDAR point clouds to create precise labels that help perception systems understand objects, people, terrain, and surrounding environments. You’ll work in specialized annotation tools, following detailed specifications while balancing accuracy, consistency, and production volume.
Your Impact
- Annotate 2D and 3D objects across images, video, and LiDAR point clouds, including vehicles, machinery, people, signage, and environmental features.
- Create and maintain accurate segmentation, object tracking, pose, and keypoint annotations across sequences.
- Apply detailed annotation guidelines, including class definitions, occlusion rules, object thresholds, and inclusion/exclusion criteria.
- Review your work, respond to QA feedback, and maintain established quality and accuracy standards.
- Identify and escalate ambiguous scenes, sensor artifacts, tooling issues, and gaps in specifications rather than making assumptions.
- Document data quality and tooling issues with clear details and reproduction steps.
- Participate in calibration sessions, guideline reviews, and team standups to maintain consistent interpretations.
- Share observations that help improve annotation guidelines, taxonomies, and edge-case documentation.
Success in this role means producing accurate, consistent annotations at the expected volume while helping maintain a reliable, high-quality dataset.
What You Bring
- Strong attention to detail and the ability to maintain accuracy while performing repetitive, visually intensive work for extended periods.
- Strong spatial reasoning skills, including the ability to interpret 3D environments and understand object size, position, and orientation.
- Ability to learn and navigate complex, purpose-built software and become productive with keyboard shortcuts and other tooling features.
- Experience working in a fast-paced, scaled environment with defined productivity, quality, or accuracy targets.
- Experience with image annotation GenAI workflows
- Reliable high-speed internet, a distraction-free workspace, and the ability to work on a company-provisioned machine within a controlled environment.
Nice to Haves
- Associate or bachelor’s degree.
- Experience with image, video, LiDAR, or 3D annotation tools.
- Experience with subsurface utility engineering (SUE), construction surveying, surveying, GIS, or other work involving spatial interpretation of physical environments.
- Experience with AutoCAD or similar type tools
- Experience in construction, mining, agriculture, industrial operations, or heavy equipment environments.
- Familiarity with point cloud data, sensor fusion, or camera-LiDAR calibration.
- Experience working with autonomous vehicles, robotics, or machine perception datasets.
Compensation
Additional Information
Why You’ll Love Working Here
At Appen, we foster a culture of innovation, collaboration, and excellence. We value curiosity, accountability, and a commitment to delivering the highest-quality AI solutions for frontier models.
You’ll work on complex challenges that shape the future of AI across industries and geographies, alongside talented people in a culture that values humility over ego. You’ll have the flexibility to deliver in a way that works for you and your team, supported by tools, resources and development opportunities to continue to build your capability over time.
About Appen
Appen has been a leader in AI training data for over 30 years. We specialise in human generated data to train, fine tune, and evaluate models across generative AI, large language models, computer vision, and speech recognition. Our AI assisted data annotation platform and global crowd of more than 1 million contributors in over 200 countries support model pre-training, supervised fine tuning, evaluation and benchmarking, safety and red teaming, and multilingual global expansion.