Onkar Gaikwad

Motivated Information Technology graduate currently pursuing M.Tech at NIT Surathkal, with a strong foundation in Machine Learning, Deep Learning, and software development. Seeking an industry-oriented role where I can apply my technical skills to solve real-world problems and contribute to scalable, data-driven systems.

Crowd Management System

• Designed a computer vision–based system to automatically detect, count, and track people in CCTV footage for real-time crowd density monitoring.
• Used YOLO deep learning models for accurate person detection and integrated tracking to analyze crowd flow and movement patterns.
• Implemented overcrowding alerts based on dynamic crowd thresholds to support public safety and event management.
• Technologies: Python, YOLO, OpenCV, Deep Learning

Smart Parking System Using A* Algorithm

• Developed an intelligent parking guidance system to direct vehicles to optimal parking slots within large parking areas.
• Implemented the A* path-planning algorithm with heuristic optimization to minimize travel distance, time, and congestion.
• Designed the system for scalability and real-time decision making, targeting smart city applications.
• Technologies: Python, A* Algorithm, Heuristic Functions

Human Action Recognition

• Built a video-based human action recognition system using a CNN–LSTM architecture to capture spatial and temporal features.
• Trained and evaluated the model on Weizmann and KTH benchmark datasets for actions such as walking, running, boxing, and waving.
• Performed frame extraction, preprocessing, and sequence modeling, achieving ~90% accuracy on Weizmann and ~85% on KTH datasets.
• Technologies: Python, OpenCV, TensorFlow

• Strong understanding of Machine Learning and Deep Learning fundamentals.
• Hands-on experience with real-time computer vision applications.
• Familiar with algorithmic problem-solving and mathematical modeling.