Rahma Sleam

Key Skills

Software

Other

OpenCV AI Kit (OAK)

Top Subject Matter

Sentiment Analysis

Social Media

Disaster Classification

Top Data Types

Text

Image

Top Task Types

Classification

Diagnosis

Freelancer Overview

My expertise in AI training data centers on comprehensive data preprocessing, exploratory data analysis (EDA), and data pipeline automation for advanced machine learning models. Through various capstone projects, I have systematically prepared datasets for neural networks by addressing missing values and outliers, specifically for Convolutional Neural Networks (CNNs) applied to medical image analysis , and preprocessing text data for LSTM-based sentiment analysis. I also developed an automated EDA tool in Python to streamline data pre-processing and interpretation. Furthermore, my background includes practical experience in extracting and structuring raw data into usable training formats. A key example is my development of a National ID Card Recognition system, where I utilized OpenCV for image preprocessing to improve text recognition accuracy before extracting the data with Tesseract OCR. This combination of image processing and text preprocessing—such as cleaning social media data for NLP disaster classification —ensures that the datasets fed into my models are clean, structured, and optimized for accurate training.

Labeling Experience

ID Card Image Annotation and Text Labeling

Opencv AI Kit OakImageClassification

Developed a National ID Card Recognition system utilizing image processing and OCR. Labeled and extracted information from images of national ID cards using Tesseract OCR and OpenCV. Improved recognition accuracy through preprocessing and ground-truth label verification. • Preprocessed ID card images for optimal OCR performance. • Labeled key fields on ID card images for AI extraction models. • Verified OCR outputs against labeled text for quality control. • Integrated labeled datasets with computer vision pipelines.

Developed a National ID Card Recognition system utilizing image processing and OCR. Labeled and extracted information from images of national ID cards using Tesseract OCR and OpenCV. Improved recognition accuracy through preprocessing and ground-truth label verification. • Preprocessed ID card images for optimal OCR performance. • Labeled key fields on ID card images for AI extraction models. • Verified OCR outputs against labeled text for quality control. • Integrated labeled datasets with computer vision pipelines.

2023 - 2023

Medical Image Annotation and AI Model Training

OtherImageDiagnosis

Built a Convolutional Neural Network for medical image analysis, focusing on disease detection. Performed exploratory data analysis and labeled images such as X-rays for supervised AI training. Addressed outliers and ensured high-quality labeled data for model development. • Annotated medical images for disease presence or absence. • Curated datasets for model training and validation. • Enhanced label consistency through review and correction. • Used CNNs to automate medical image diagnosis tasks.

Built a Convolutional Neural Network for medical image analysis, focusing on disease detection. Performed exploratory data analysis and labeled images such as X-rays for supervised AI training. Addressed outliers and ensured high-quality labeled data for model development. • Annotated medical images for disease presence or absence. • Curated datasets for model training and validation. • Enhanced label consistency through review and correction. • Used CNNs to automate medical image diagnosis tasks.

2023 - 2023

NLP-Based Disaster Tweet Data Labeling

OtherTextClassification

Conducted NLP-based classification of disaster-related tweets leveraging machine learning. Preprocessed and manually labeled tweet data to train and validate classification models. Enhanced machine intelligence for disaster response through labeled social media data. • Developed and iterated on disaster classification labels for text data. • Utilized NLP toolkits to preprocess and annotate tweets. • Evaluated ML model predictions and iteratively improved label quality. • Collaborated on labeling guidelines for effective disaster identification.

Conducted NLP-based classification of disaster-related tweets leveraging machine learning. Preprocessed and manually labeled tweet data to train and validate classification models. Enhanced machine intelligence for disaster response through labeled social media data. • Developed and iterated on disaster classification labels for text data. • Utilized NLP toolkits to preprocess and annotate tweets. • Evaluated ML model predictions and iteratively improved label quality. • Collaborated on labeling guidelines for effective disaster identification.

2023 - 2023

Sentiment Analysis Data Labeling and AI Training

OtherTextClassification

Developed an LSTM model for sentiment analysis on text data, including social media posts and reviews. Performed exploratory data analysis on datasets such as Sentiment140 and Amazon Reviews. Conducted feature analysis and addressed data quality issues for improved classification performance. • Preprocessed and labeled text data for sentiment categories. • Built, trained, and validated LSTM models on labeled datasets. • Applied NLP techniques for enhanced classification accuracy. • Managed and analyzed large-scale text data for AI training.

Developed an LSTM model for sentiment analysis on text data, including social media posts and reviews. Performed exploratory data analysis on datasets such as Sentiment140 and Amazon Reviews. Conducted feature analysis and addressed data quality issues for improved classification performance. • Preprocessed and labeled text data for sentiment categories. • Built, trained, and validated LSTM models on labeled datasets. • Applied NLP techniques for enhanced classification accuracy. • Managed and analyzed large-scale text data for AI training.

2023 - 2023

Education

Work History