DragLord 341

Developed and compared five deep learning models for speech emotion recognition on the EMO-DB dataset, achieving a maximum test accuracy of 80.54% with the CNN-based AlexNet model. Engineered a robust audio preprocessing pipeline leveraging OpenSMILE and Librosa, and optimized convergence using PyTorch-WideDeep with Xavier Normal initialization and Adam optimizer.

Led development of an AI-driven robotics navigation system, integrating AI-based image recognition, STM32 microcontroller commands, and Bluetooth communication. Spearheaded the creation of a distributed system for real-time communication between devices, cutting down latency by 15% and improving overall system performance by 30%.

Pioneered development of a React-based frontend for a dynamic website, building over 10 different webpages with user registration and account creation capabilities. Innovated by implementing a central Firebase database and integrating six different APIs to provide real-time BTO property details and nearby amenities.

Completed a comprehensive web development foundations course, acquiring proficiency in HTML, CSS, and JavaScript. Developed hands-on projects, including a Recipes webpage, to reinforce practical application and problem-solving skills within a virtual learning environment.

Led a team of six in designing and implementing a comprehensive camp management system using Java in Visual Studio Code. Developed three distinct user roles with customized permissions and created Javadoc documentation to ensure code readability and maintainability.

Conducted in-depth data analysis using Jupyter Notebook to explore correlations in a 100k+ record dataset, applying Machine Learning Models like Catboost, Q-Learning, and Neural Networks to achieve over 80% accuracy. Utilized Optuna for hyperparameter optimization, increasing model accuracy by 15%, and employed F1-score and correlation matrices for robust evaluation and feature selection.

Grade: 4.10

Activities and societies: Machine Learning Project: Speech Emotion Recognition, Robotics Navigation System: AI-Driven Pathfinding Project, Full Stack Application: BTOgether, Machine Learning Project: Drug Usage Correlations

Relevant coursework included Neural Networks and Deep Learning, Intelligent Agents, Cloud Computing, Machine Learning, Data Structures and Algorithms, Artificial Intelligence, Database, Operating Systems, and Object-Oriented Programming. Achieved a GPA of 4.10.