Python developer with experience in building web applications and a passion for data science. Currently learning data science techniques and applying them to real-world problems.
- Proficient in Python and related web development technologies (e.g. Flask, Django) π
- Familiar with data science tools and libraries such as NumPy, Pandas, and Scikit-learn π
- Strong understanding of data structures and algorithms π§©
- Experience with SQL and database management ποΈ
- π» Automatic Number-Plate Recognition (ANPR) using YOLOv4: The model localises the Licence Plate and crops it and sends it to do OCR. I have used YOLOv4 tiny and trained on a total of 2000 images for 6000 epochs. It has an accuracy of about 97 Percent on the test set.
- π€ Head Pose Estimation using Mediapipe: This application is developed using Javascript. It used Mediapipeβs FaceMesh to get the Facial Keypoints and converts them into points in 3D space.
- ποΈ Motorcycle Helmet Detection using YOLOv4: The model detects Helmet, Non-Helmet and the entire rider+bike. If we find a person not wearing a helmet inside the entire rider+bike bounding box, the bike is sent for ANPR.
- π Deep Reinforcement Learning by Udacity, 2020.
- π Machine Learning by Stanford University, 2020.
- π§ [email protected]
- π₯ https://www.linkedin.com/in/ebin-babu-thomas-06942a17b/
- π¦ @ebin005