Ashley Robinson

Key Skills

Software

CVAT

Labelbox

Scale AI

Clickworker

Top Subject Matter

Autonomous Vehicle Imagery

Multilingual NLP Annotation

Computer Vision for Object Detection

Top Data Types

3D Sensor

Computer Code Programming

Text

Top Task Types

Bounding Box

Computer Programming Coding

RLHF

Text Generation

Text Summarization

Freelancer Overview

I'm an experienced AI Data Labeling and NLP Specialist with a passion for helping build the next generation of autonomous systems and Large Language Models (LLMs). Over the years, I've honed my skills in data annotation, particularly in computer vision tasks like traffic sign recognition and object detection, as well as in multilingual NLP. I take pride in my ability to deliver high-quality, accurate training data that truly makes a difference in AI development. I've had the opportunity to work on exciting projects, from improving autonomous vehicle systems to enhancing language models that understand multiple languages. I’m also skilled at using tools like Labelbox and Roboflow, and I enjoy leading teams and sharing my knowledge with others. My focus is always on getting the details right while ensuring that projects are completed efficiently and effectively.

Labeling Experience

Image Annotation

CVATImageBounding BoxPolygon

Title: NBA Basketball Players Annotation This project involved annotating images of NBA basketball players. The primary goal was to accurately label player positions and movements using multiple annotation methods including bounding boxes, polygons, and segmentation. The project was executed using CVAT software to ensure precise labeling of complex images. The annotations were used to train machine learning models for sports analytics and player tracking applications. The project encompassed thousands of images and focused on maintaining a high standard of quality, with regular reviews to ensure adherence to accuracy metrics such as IoU (Intersection over Union) and manual quality control. The labeling tasks were performed with a particular focus on the consistency and relevance of the annotations to the end-use of the dataset.

Title: NBA Basketball Players Annotation This project involved annotating images of NBA basketball players. The primary goal was to accurately label player positions and movements using multiple annotation methods including bounding boxes, polygons, and segmentation. The project was executed using CVAT software to ensure precise labeling of complex images. The annotations were used to train machine learning models for sports analytics and player tracking applications. The project encompassed thousands of images and focused on maintaining a high standard of quality, with regular reviews to ensure adherence to accuracy metrics such as IoU (Intersection over Union) and manual quality control. The labeling tasks were performed with a particular focus on the consistency and relevance of the annotations to the end-use of the dataset.

2022 - 2024

AI Code Review Task

ClickworkerComputer Code ProgrammingComputer Programming Coding

The project involved reviewing AI-generated code, ensuring that the outputs adhered to project-specific guidelines, and maintaining high standards of code quality. Tasks included evaluating code functionality, identifying bugs, and verifying the correctness of code snippets in various programming languages. The project was conducted at Scale AI, and strict quality measures were implemented, ensuring compliance with coding best practices and performance metrics. The scope of the project covered large-scale datasets, with a focus on improving the accuracy and efficiency of AI coding models.

The project involved reviewing AI-generated code, ensuring that the outputs adhered to project-specific guidelines, and maintaining high standards of code quality. Tasks included evaluating code functionality, identifying bugs, and verifying the correctness of code snippets in various programming languages. The project was conducted at Scale AI, and strict quality measures were implemented, ensuring compliance with coding best practices and performance metrics. The scope of the project covered large-scale datasets, with a focus on improving the accuracy and efficiency of AI coding models.

2021 - 2021

Computer Vision Data Labeling Expert

LabelboxVideoBounding BoxPolygon

The project involved extensive data labeling tasks across images and videos for computer vision models. Specific tasks included drawing bounding boxes, segmenting objects using polygons, and annotating text using Named Entity Recognition (NER) methods. The project was performed using CVAT, Labelbox, and Scale AI tools, adhering to strict quality guidelines to ensure precise labeling for use in autonomous vehicle development, healthcare diagnostics, and large language models (LLMs). The scope of the project covered a large dataset with a focus on accuracy, efficiency, and scalability of the labeling process.

The project involved extensive data labeling tasks across images and videos for computer vision models. Specific tasks included drawing bounding boxes, segmenting objects using polygons, and annotating text using Named Entity Recognition (NER) methods. The project was performed using CVAT, Labelbox, and Scale AI tools, adhering to strict quality guidelines to ensure precise labeling for use in autonomous vehicle development, healthcare diagnostics, and large language models (LLMs). The scope of the project covered a large dataset with a focus on accuracy, efficiency, and scalability of the labeling process.

2020 - 2020

Education

Work History