I'm a developer with a focus on AI/ML, computer vision, and full-stack development. Below is a brief overview of my projects, each showcasing different technologies and frameworks in action.

Linkedin https://www.linkedin.com/in/harshil-c/

1. FaceFit - AR-Based Virtual Accessory Try-On

   A cutting-edge augmented reality application that allows users to try on virtual accessories in real-time. This project focuses on delivering smooth performance and an intuitive user interface, using advanced face detection and AR technologies to provide a seamless experience.

   Tech: Java Spring Boot, Mediapipe.js for real-time AR effects, PostgreSQL for data management

2. 3D Image Composer

   An interactive tool designed for creating anaglyph 3D images by segmenting a person from an image, integrating them into stereoscopic backgrounds, and enabling dynamic depth adjustment. Users can fine-tune positioning and visualize depth using red-cyan 3D glasses, with a clean interface powered by Gradio.

   Tech: Python, TensorFlow for image segmentation, Gradio for UI, NumPy for computations, Matplotlib for visualizations

3. Vision Transformers ViT DEMO

   Demonstrates the potential of Vision Transformers (ViT) in image classification tasks. This project highlights the effectiveness of transformer-based architectures in computer vision and includes comprehensive visualizations for better understanding.

   Tech: Python, PyTorch for model implementation, Matplotlib for data visualization

4. Building Elevator Model

   A simulation of a multi-elevator system in a building, managing concurrent elevator requests and updates in real-time. Built using MVC architecture, the system is multithreaded for asynchronous control and features a functional graphical user interface.

   Tech: Java, MVC pattern, Swing for GUI, multithreading for elevator logic

5. IMC Competition Prosperity 2

   Developed an algorithmic trading model for the IMC Trading International Hackathon. This project utilized real-time data analysis, quantitative techniques, and linear regression to predict trading trends, achieving a top 1% global ranking.

   Tech: Python, Pandas for data manipulation, NumPy for numerical computations, Matplotlib for visualization

6. Stock Price Prediction and Analysis

   A data-driven project analyzing historical stock data to predict future trends. The machine learning pipeline includes preprocessing, feature engineering, and training predictive models to provide actionable insights, with detailed visualizations for trend analysis.

   Tech: Python, Scikit-Learn for machine learning, Pandas for data processing, Matplotlib and Seaborn for visualization

• Machine Learning: PyTorch, Transformers, Scikit-Learn, NumPy, Pandas
• Computer Vision: OpenCV, Mediapipe.js, TensorFlow, ViT, Swin
• Web Development: Java, Spring Boot, Flask, Express (Node.js)
• Databases: PostgreSQL
• Visualization: Matplotlib, Seaborn