Rakibul

Introduction

In the different realms of science and technology, it is hard to believe that an Artificial Intelligence(AI) domain is going to change the world as much as advanced software will in the decades ahead. As Machine Learning, Deep Learning and Computer Vision models evolve, they will become more capable of handling more complex tasks and offering more efficient solutions.

We have already witnessed the AI's success. In addition, many AI models struggle when the data changes, often failing to apply what they've learned in training to real-world situations. A clear example of this is how AI systems frequently have trouble recognising data from underrepresented groups. Another important issue with these models is how efficiently they work, especially now that AI is being used in so many areas.

I'm drawn to solving real-world problems and enjoy exploring how AI can be applied across different fields. I'm passionate about machine learning, deep learning, and computer vision—especially when it comes to building trustworthy and efficient AI systems that can tackle real-world problems.

Research Focus

As we collect more data and use AI in sensitive areas like healthcare, education, and autonomous systems, it's more important than ever to build intelligent systems that focus on people, work efficiently, need less data, and can clearly explain their decisions. To help move this vision forward, I plan to tackle these challenges from four key angles:

1. Human-Centred AI System Design

My goal is to design AI systems and large-scale databases that enhance human problem-solving rather than replace it. I want to create tools that support people, integrating domain knowledge and user-friendly interfaces, so the systems work alongside humans and help them perform better.

2. Computational Efficiency

Many cutting-edge deep learning models demand huge amounts of computational power, which makes them hard to use in real-time or resource-limited environments. I'm especially interested in developing lightweight models, compression techniques, and optimization strategies that make AI more efficient and easier to deploy.

3. Data-Efficient Learning

Most deep learning approaches depend heavily on large, annotated datasets, which are costly and time-consuming to create. I want to focus on methods like few-shot learning, semi-supervised learning, and transfer learning to build models that can still perform well with limited data, making AI more accessible and fair.

4. Explainable and Trustworthy AI

As AI becomes part of critical decision-making systems, we must understand how these systems work. I aim to build explainable AI frameworks that communicate how predictions are made, especially for non-experts, so people can trust, use, and regulate AI systems responsibly.

Research Experience

One of the most memorable and rewarding experiences during my undergraduate studies was taking a class on Machine Learning. The course was well-structured and completely changed how I think about problem-solving it showed me that there are countless ways to approach a challenge using the right tools and mindset. Since then, I've been especially interested in applying AI techniques to help systems interact with the real world and make smarter, more effective decisions.

I am passionate about building machine-intelligent systems that process information at large-scale data and assist humans in various knowledge-intensive tasks.

Thesis Research: Deep Learning for Audio Enhancement

The thesis experiences gave me a deeper understanding of practical research problems, enhanced my programming and software skills, and broadened my knowledge of deep learning and operating environments. I research deep learning-based background noise classification and reduction for audio enhancement. In this research, we use a sliding window to constantly run over the background sound wave and identify the background sound. This process will also reduce the background noise from our focused sound. In future work, it can be implemented in real-time virtual conversations.

In this thesis, we overcame each challenge through rapid prototyping, leveraging interdisciplinary techniques from psychology, and fostering effective team communication, all driven by my research principles and passion. Interacting with my supervisor on my thesis, experiencing an intellectual atmosphere at my institute, has further motivated me to pursue a research career.

Future Goals

I aspire to spend the next several years pursuing graduate studies (MS/PhD) in computer science with a focus on machine learning, deep learning, computer vision and AI systems. During this time, I hope to gain rigorous theoretical grounding, conduct impactful research, and collaborate across disciplines.

My long-term goal is to become a researcher who advances responsible and efficient AI, developing models that are not only accurate but also usable, ethical, and accessible to all.