Key Skills

Software

CVAT

Top Subject Matter

No subject matter listed

Top Data Types

Image

Top Label Types

Bounding Box

Segmentation

Object Detection

Action Recognition

Prompt Response Writing SFT

Freelancer Overview

I am a data annotation specialist with over 4 years of hands-on experience supporting computer vision and machine learning projects at leading AI organizations. My expertise spans image, video, and 3D point cloud annotation using industry-standard tools such as CVAT, Label Studio, and Labelbox. I have consistently maintained 98%+ accuracy rates across large-scale datasets for applications including autonomous vehicles, robotics, medical imaging, and surveillance, leveraging advanced techniques like semantic and instance segmentation, keypoint detection, and object tracking. I excel at collaborating with data scientists and engineers to refine guidelines, conduct thorough quality assurance, and optimize annotation processes. My strong technical background in computer vision, meticulous attention to detail, and ability to adapt to complex requirements enable me to deliver high-quality training data that drives reliable AI model performance.

Labeling Experience

Geospatial & Object Detection Annotation for Autonomous Vehicle Datasets

CVATImageBounding BoxSegmentation

Executed high-precision annotation on a large-scale dataset consisting of 50,000+ urban and rural images intended for autonomous vehicle navigation. Tasks included 2D bounding box placement for vehicles, pedestrians, and signage, as well as 3D cuboid fitting for depth perception. Adhered to strict NU (Null) and IO (Ignore) guidelines to ensure edge cases were handled correctly. Achieved a 98% inter-annotator agreement rate by consistently following dynamic QA feedback loops and utilizing "supervision" mode for corner-case scenarios

Executed high-precision annotation on a large-scale dataset consisting of 50,000+ urban and rural images intended for autonomous vehicle navigation. Tasks included 2D bounding box placement for vehicles, pedestrians, and signage, as well as 3D cuboid fitting for depth perception. Adhered to strict NU (Null) and IO (Ignore) guidelines to ensure edge cases were handled correctly. Achieved a 98% inter-annotator agreement rate by consistently following dynamic QA feedback loops and utilizing "supervision" mode for corner-case scenarios

2022 - 2023

Education

Work History