Ndiaye Joelmamadou

Key Skills

Software

Img Lab

LabelImg

Remotasks

SuperAnnotate

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Classification

Object Detection

Question Answering

Tracking

Freelancer Overview

In the field of data annotation and AI training data, I have solid practical experience and the ability to adapt to multiple scenarios, especially excelling in image annotation and quality control. My core skills include precise target positioning (pixel-level boundary sketching, multi-category object classification), application of mainstream tools (LabelMe/CVAT/Labelbox, etc.), and the implementation of cross-industry annotation standards. I can efficiently handle annotation tasks in scenarios such as industrial quality inspection, retail product recognition, and security monitoring. I have deeply participated in a certain intelligent security project, leading the annotation of over 100,000 frames of video images, accurately identifying 8 types of targets (pedestrians / vehicles / suspicious objects, etc.), with an annotation accuracy rate of over 97%; I completed 50,000+ image classification annotations for e-commerce platforms, supporting the optimization of image search functions, and simultaneously establishing a "common target feature library", helping the team increase daily efficiency by 20%. I am familiar with the entire process from single-category annotation to multi-target processing in complex scenarios, emphasizing consistency and traceability of annotation standards, and can quickly respond to the customized requirements of different projects, providing high-quality data support for AI model training.

Labeling Experience

Industrial Part Defect Detection Image Annotation Project

LabelimgImageBounding Box

This project focuses on the annotation of part images in industrial manufacturing. Using the CVAT software, the boundaries of defects such as scratches, cracks, and holes on the part surfaces are marked. A total of over 20,000 + images have been processed. During the execution, strict quality control measures were strictly followed. After multiple rounds of cross-validation, the annotation accuracy rate exceeded 98%, providing high-quality data support for the training of AI models in the industrial quality inspection field, and helping to improve the efficiency and accuracy of automated quality inspection.

This project focuses on the annotation of part images in industrial manufacturing. Using the CVAT software, the boundaries of defects such as scratches, cracks, and holes on the part surfaces are marked. A total of over 20,000 + images have been processed. During the execution, strict quality control measures were strictly followed. After multiple rounds of cross-validation, the annotation accuracy rate exceeded 98%, providing high-quality data support for the training of AI models in the industrial quality inspection field, and helping to improve the efficiency and accuracy of automated quality inspection.

2017 - 2000

Education

Work History