Amit Verma

Results-driven Computer Science student specializing in Machine Learning (PyTorch), Deep Learning, Computer Vision, and Image De-noising. Skilled in AI research, backend development, and full-stack development with strong problem-solving skills.

AdaptiveSpatialNorm

February 2025

Developed an Adaptive Spatial Normalization method to dynamically normalize spatial features, enhancing model accuracy in varying data conditions by prioritizing key features and normalizing the image relative to them.

Normalization Comparison

February 2025

The paper compares traditional and modern De-noising models, evaluating their performance, accuracy, and efficiency across different noise types.

• Auto Routine (Technologies: JavaScript, HTML, CSS, Python) – Generates an automated routine using the number of teachers and classes.
• Snake Game (Technologies: HTML, CSS,JavaScript) – Demonstrates pathfinding and AI-driven gameplay strategies.
• Neural Network in C++ (Technologies: C++) – Built a neural network from scratch for efficient computation.
• Cursor Movement and Volume Control (Technologies: Python, OpenCV, MediaPipe) – Implemented hand-tracking to control system volume.
• Weather Prediction (LSTM) (Technologies: Python, LSTM, PyTorch) – Designed a weather model using LSTM.
• Real-Time Stock Prediction (Technologies: Python, LSTM, yfinance, PyTorch) – Developed a real-time stock prediction model using live data.
• SSPL App (Technologies: Python, Qt Designer, OpenCV, NumPy, MATLAB) – Developed an app with filters, edge detectors, and video customization.