Tharun Aakash

Key Skills

Software

Roboflow

Top Subject Matter

No subject matter listed

Top Data Types

Image

Top Label Types

Bounding Box

Polygon

Point Key Point

Segmentation

Freelancer Overview

I am a Computer Science student specializing in AI-driven software development, with hands-on experience in building intelligent systems that rely heavily on high-quality data annotation and training data pipelines. My work includes developing a CNN-based distracted driver detection platform and AI-powered medical image analysis tools, where I managed the full cycle of data collection, preprocessing, labeling, and model training using Python, TensorFlow, Keras, OpenCV, and scikit-learn. I am skilled in handling large datasets, optimizing data pipelines, and ensuring data integrity for robust model performance in computer vision and NLP domains. My technical background also includes designing scalable backend systems and integrating AI workflows into production environments, which has given me a strong foundation in both the practical and technical aspects of creating and managing training datasets for machine learning applications.

Labeling Experience

Jewel AI

RoboflowImageBounding BoxPolygon

The Jewel Try-On System is an AI-driven virtual try-on solution that realistically places earrings, rings, and necklaces on user photos using computer vision and deep learning. It combines pose estimation, segmentation, and physics-based scaling to ensure accurate size, placement, and orientation. Advanced lighting, shadow, and occlusion modeling produce photorealistic results suitable for e-commerce and virtual styling applications.

The Jewel Try-On System is an AI-driven virtual try-on solution that realistically places earrings, rings, and necklaces on user photos using computer vision and deep learning. It combines pose estimation, segmentation, and physics-based scaling to ensure accurate size, placement, and orientation. Advanced lighting, shadow, and occlusion modeling produce photorealistic results suitable for e-commerce and virtual styling applications.

2025 - 2025

Education

Work History