Shazzadul Islam - Course & Project

Name: ChronoTick (Benchmarking LLM Performance vs Accuracy)

Course: Advanced Operating Systems (CS550)Supervisor: Dr. Sun, Xian-HeLanguage: Python, PyTorch, Time-Series Forecasting Models (Chronos, TimesFM, Moirai, Time-MoE)Specification: ChronoTick is a software-defined clock that leverages advanced time-series forecasting models to maintain an up-to-date view of the physical clock. Inspired by AI-based frame generation, it employs “timegen” to proactively generate clock updates, informed by real-time node conditions (e.g., temperature, power). The system integrates a feedback loop with sparse NTP/PTP corrections to prevent drift, enabling fine-grained synchronization across distributed nodes. Contributions include: (i) a trade-off study of forecasting models for wall-clock prediction, (ii) the ChronoTick design for continuous proactive updates, and (iii) demonstrated improvements in synchronization for multi-node algorithms. [Presentation]

Name: Real-Time ML data streaming pipeline

Course: Cloud Computing (CS553)Supervisor: Dr. Sun, Xian-HeLanguage: Python, Apache Flink, Jupyter Notebook, BentoMLSpecification: The ML Data Stream Pipeline project focuses on building a real-time data processing framework capable of ingesting and analyzing continuous data streams. It integrates Apache Flink for stream processing and BentoML for serving ML models, enabling both text-based (word count) and image-based (decoding) pipelines. The system demonstrates how machine learning inference can be applied to live data streams with a modular, scalable, and low-latency design. [Git]

North South University

Name: Smart Street: Street of Next Generation

Course: Capstone Design (CSE499A and 499B)Supervisor: Mr Fahimul HaqueLanguage: Python, C++, Flask, ArduinoSpecification: The Smart Street system features various robust functionalities, including Intelligent Lighting, Smart Parking, Smart Traffic Control, and Safe Zebra Crossing. These features aim to address various issues related to multiple aspects of road safety and traffic rules, utilizing IoT and machine learning approaches. For the Machine learning part, we have to collect traffic data manually from multiple locations in Bangladesh. [poster] [video] [Git]

Name: Detecting Fake Bengali Names From Social Networks

Course: Directed Research (CSE498R) Supervisor: A K M Bahalul HaqueLanguage: PythonSpecification: In this project, we explore natural language processing techniques for the detection of a fake name in the context of Bangladesh. To extract name features, we use Term Frequency-Inverse Document Frequency and N-grams of letters. After getting vector features, we feed those models into the ML pipeline. The main techniques are supervised learning pipelines to classify names that do not represent real Bengali names. We ran different Machine Learning algorithms such as Naïve Bayes, Random Forest (RF), and Support Vector Machine (SVM).

Name: Language Identification from Human Voice

Course: Pattern recognition and neural network(CSE465)Supervisor: Dr. Mohammad Ashrafuzzaman Khan Language: PythonSpecification: The project goal was to identify four different languages, including Bengali, from human voices using a deep learning approach. We selected frequency and time to get the feed in the form of a spectrogram. Using Scale-invariant Feature Transform Features on spectrogram images of speech signal combination with SVM and LSTM that classify different languages.[Git]

Name: Bird Species Classification from an Image Using VGG-16 Network

Course: Machine Learning (CSE445) Supervisor: Khan Mohammed HabibullahLanguage: PythonSpecification: In this project, the main objective is to classify Bangladeshi birds according to their own species using several machine learning algorithms through transfer learning. The dataset for this classification is one that I collected manually and consists only of bird species that are found in Bangladesh. This was done because there is no collection of local bird data in Bangladesh. We used the VGG-16 network as our model to extract the features from bird images. To perform the classification, we used several Algorithms. However, compared to other classification methods, such as Random Forest and K-Nearest Neighbor (KNN), the Support Vector Machine (SVM) achieved the maximum accuracy of 89%. [paper]

Name: Gryffindor Education web(Extended CSE 299)

Course: Internet and Web Technology (CSE482)Supervisor: Md Naqib Imtiaz HussainLanguage: PHP, HTML, CSS, Bootstrap, JavaScript, MySQL Specification: This project is an extended version of the junior design project. The core goal of this project is to create a robust learning experience that mimics a traditional classroom experience, similar to Google Classroom. A student can enroll in a course and see the course material. Teachers can take quizzes and upload assignments. Student can post their problem and view video material. The main feature of this project is the live class. The teacher can take live classes when needed. [Demo] [Git]

Name: Gryffindor Education web.

Course: Junior Design (CSE299) Supervisor: Dr Shahnewaz SiddiqueLanguage: PHP, HTML, CSS, Bootstrap, MySQL Specification: The core goal of this project is to create a robust learning experience that feels like a classroom experience similar to Google Classroom.A student can enroll in a course and see the course material. Teachers can take quizzes and upload assignments. Student can post their problem and view video material. [Demo] [Specification] [Git]

Name: E-Learning platform, both Android and web.

Course: Software Engineering (CSE327) Supervisor: AKM Bahalul HaqueLanguage: PHP, HTML, CSS, Bootstrap,spring-boot, JAVA, MySQL Specification: E-learning is an integral part of innovative education. Many e-learning systems are widely available to educational institutions. The challenge is to easily integrate the e-learning system into an innovative educational environment that meets the users' requirements. [Git]

Name: Object Detection using Raspberry pie

Course: Operating Systems Design (CSE323) Supervisor: AKM Bahalul HaqueLanguage: PythonSpecification: The main idea of the project is to deploy an OpenCV object detection pipeline on a Raspberry Pi to detect objects. We use a Raspberry Pi 3 and a Pi Camera as our hardware module. A Python script runs on the Raspbian OS and starts streaming video. Those video frames feed into the detection pipeline and show results simultaneously.[Specification]

Name: Design and Implementation of a 16-bit single-cycle data path design

Course: Computer Organization and Architecture (CSE332) Supervisor: Shaikh Shawon-Arefin-ShimonLanguage: Logisim Specification: Our task was to design an 8-bit single-cycle CPU that has separate Data and Instruction Memory. It should also be able to connect with at least two output devices. We use an accumulator-based architecture, where we will write and read data on the accumulator. This accumulator will be connected to the register file to facilitate data movement between the registers and the accumulator, and vice versa. One Operand. Operands are register-based. And saved to or fetched from the accumulator $acc. [Specification]

Name: Online Restaurant Management System

Course: Database Systems (CSE311) Supervisor: Khan Mohammad HabibullahLanguage: PHP, HTML, CSS, MySQL Specification: This project is a web-based application, utilizing raw HTML, CSS, and PHP. This project has two sides: admin and user. Admin can add food and also update food. Take online orders and deliver food to the customer. Users can search for food or a restaurant and add food to a cart. Finally, they can order their selected food from this website.[Git]

Name: Design and Implementation of a Particular Text with a Combinational and Sequential Circuit

Course: Digital Logic Design (CSE231) Supervisor: Dr. Arshad M Chowdhury Language: Logisim Specification: The main idea of the project is to deploy a total circuit that will automatically show CSE231-1-11 in a single 7-segment display. It can be in both forward and reverse (11-1-132ESC) ways. The total design of this circuit will be designed in three phases. In the first phase, we will design the combinational part of the circuit, which involves specifying our inputs and determining the outputs of the circuit, and then proceed to the formulation step. After formulating them, we will further implement the circuit with our theoretically simplified design, and finally check and verify if we are getting all the expected outputs for the inputs as defined in the “specification” step, as well as our desired outputs. Later, we will design the sequential circuit with the facilitation of forward and reverse sequences. Further, the sequence will be implemented according to the prior state diagram design, according to the procedure.

Name: Weight Balance Tree

Course: Data Structure and Algorithms(CSE225)Supervisor: Ms. Tamanna Motahar Language: C++Specification:

Name: Advising system

Course: Introduction to Object-Oriented Programming (CSE215) Supervisor: Mr. Adnan Firoze Language: JAVA, SQLiteSpecification: This project was focused on Advising and student profiles, especially for the students. This Swing GUI-based interface has the support of saving and retrieving data using a SQLite Database. Student can advise their courses. They can see their CGPA and ID. They can add courses and also delete classes. This system also calculates tuition fees.

Name: Library Management System

Course: Programming Language 1(CSE115)Supervisor: Rashed Selim TopuLanguage: CSpecification: