Enoch A.

Key Skills

Software

Roboflow

LabelImg

Label Studio

Top Subject Matter

Agriculture – Crop Disease Detection & Precision Farming

Technology – Computer Vision & Object Detection

Engineering – Embedded Systems & Autonomous Systems

Top Data Types

Image

Top Task Types

Bounding Box

Freelancer Overview

The most direct AI training data I have had in my own experience is with creating a custom object detection system for my final year project, an autonomous drone to determine the state of crop disease. In my case, I used Roboflow to curate and label a custom image dataset using Roboflow's Label It application, leaving high quality, consistent training data in each crop disease class. This practical session provided me with a solid insight into the principles of a good annotation, which included accurate detection boundaries, uniformity of labelling conventions, and clear distinction between different classes, as I learned firsthand how the quality of the annotations significantly influenced my YOLOv8n model's performance in detecting each class. In addition to labelling work, I have a strong background in Computer Engineering and certifications in Data Science (Cisco) and AI (Huawei ICT v4.0), which has provided me with a solid foundation in the working of machine learning pipelines from data collection to deployment. I know the intended use of each label I create and can be more accurate and consistent than a labeler without this understanding. I also have the ability to type at 50 words per minute, am very detail-oriented and enjoy working alone on structured remote tasks.

Labeling Experience

Tomato crop disease

ImageBounding Box

Cured and annotated a customized image dataset of tomato leaf diseases with Roboflow for bounding box labelling. Classed hundreds of images with multiple classes, properly placed the classes and used consistent class naming conventions for the whole dataset. A labelled dataset was used to train a YOLOv8n object detection model that was deployed on an autonomous drone equipped with a Raspberry Pi 4B to detect crop diseases in real time. Annotations were iteratively checked and corrected to maintain quality prior to export for a clean, model-ready set of data.

2026 - Present

Education

Work History