
Bristi Rani Roy
Assistant Professor
At a glance
Biography
Bristi Rani Roy is a researcher specializing in Machine Learning, Artificial Intelligence, Natural Language Processing, Image Processing, and Deep Learning. She has contributed to peer-reviewed publications on machine learning applications for breast cancer diagnosis, skin disease classification, and email spam detection, in conferences such as ICAICT 2020 and journals such as IRJET. Her research also extends to text watermarking algorithms using NLP techniques, demonstrating her expertise in both healthcare informatics and cybersecurity applications. She has completed training in instructional methods and mobile application development from Bangladesh Army University of Engineering & Technology and the ICT Ministry of Bangladesh. With diverse project experience spanning web development, Android applications, and computer graphics, she continues to advance AI-driven solutions for real-world challenges.
Education
| Degree | Institution | Year | Result |
|---|---|---|---|
| M.Sc.Engg. in Computer Science and Engineering | Rajshahi University of Engineering & Technology | 2025 | |
| B.Sc. in Computer Science and Engineering | Bangladesh Army University of Engineering & Technology | 2019 |
Research Interests
Publications
Journal Publications
Spectral–Structural Fusion for Hyperspectral Image Anomaly Detection Using a Hybrid RX and SSIM-3D CNN Fusion Model
This paper proposes a Spectral–Structural Fusion framework for hyperspectral image (HSI) anomaly detection by integrating the statistical capabilities of the Reed–Xiaoli (RX) algorithm with the structural similarity features extracted using SSIM and a 3D Convolutional Neural Network (3D-CNN). The hybrid model combines spectral information with spatial–structural characteristics to improve the detection of anomalous pixels in complex hyperspectral scenes. The proposed approach aims to achieve more accurate and robust anomaly detection while reducing the limitations of relying solely on spectral or spatial information.
Conference Publications
This paper presents an offline signature verification system based on Siamese Neural Networks, designed to determine whether two handwritten signatures belong to the same individual. The study investigates and compares four backbone architectures—MobileNetV2, VGG16, InceptionV3, and a custom CNN—within the Siamese framework. The proposed approach learns discriminative signature features and measures similarity between genuine and questioned signatures. The comparative evaluation highlights the effectiveness of different architectures for achieving reliable and robust offline signature verification.
This paper evaluates the effectiveness of 3-D DenseNet and Spatial-Spectral Residual Networks (SSRN) for hyperspectral image classification. It investigates how these deep learning architectures exploit the rich spatial and spectral information of hyperspectral images to improve classification performance. The study compares their classification accuracy and related performance measures on benchmark hyperspectral datasets, highlighting their strengths, limitations, and suitability for different classification scenarios.
Evaluating and enhancing the performance of skin disease classification based on ensemble methods
This paper evaluates and enhances the performance of skin disease classification using ensemble machine learning methods. It investigates multiple classification models and combines their predictions to improve accuracy, robustness, and generalization. The study aims to develop a reliable automated approach for distinguishing different skin diseases from medical images, supporting efficient and accurate computer-aided diagnosis.
Professional Experience
Lecturer
15 September 2019 - January 2026Dept. of Computer Science & Engineering, Bangladesh Army University of Engineering & Technology
Honors & Awards
Gold Medal
Proud to have received the Gold Medal for outstanding academic performance in my B.Sc. in Engineering. This achievement reflects my dedication, hard work, and commitment to academic excellence.
Vice Chancellor’s Award
Received the Vice Chancellor’s Award in recognition of outstanding academic performance and excellence during my B.Sc. in Engineering.
Dean Award
Honored with the Dean’s Award in every semester for consistent academic achievement and dedication to excellence during my B.Sc. in Engineering.
BAUET Merit Scholarship
BAUET Merit Scholarship
Workshops & Seminars
Workshops
Hands-on-orientation session "Outcome Based Education"
This workshop provides an overview of Outcome-Based Education (OBE), focusing on its principles, implementation, and assessment practices. It helps participants understand Course Outcomes (COs), Program Outcomes (POs), Program Educational Objectives (PEOs), CO–PO mapping, and outcome-based assessment to ensure continuous improvement in academic programs.
Outcome Based Education
This workshop provides an overview of Outcome-Based Education (OBE), focusing on its principles, implementation, and assessment practices. It helps participants understand Course Outcomes (COs), Program Outcomes (POs), Program Educational Objectives (PEOs), CO–PO mapping, and outcome-based assessment to ensure continuous improvement in academic programs.
Skill Development for Mobile Game & Application Project
Skill Development for Mobile Game & Application Project was a three-month workshop held in 2018, focused on developing practical skills in mobile game and application development. The workshop provided hands-on training in application design, programming, project development, and problem-solving, enabling participants to build and implement mobile-based projects.
Seminars & Talks
An Introspective on Statistical Signal Processing in the Age of Generative AI
A seminar titled “An Introspective on Statistical Signal Processing in the Age of Generative AI” was held on 13 August 2026 at SPICSCON 2026. The seminar was conducted by Professor Junsong Yuan, Director of Graduate Admissions, Department of Computer Science and Engineering, University at Buffalo, State University of New York. The talk explored the evolution of statistical signal processing in the era of Generative AI, focusing on applications such as speech recognition, object detection, and video analytics. It also discussed key challenges including the curse of dimensionality, overfitting, and the gap between low-level features and high-level semantics, while highlighting emerging research opportunities and the limitations of GenAI.
Training: Method of Instructions (MoI)
Other Information
Responsible for handling BAUET website
Assistant Advisor
BAUET Computer Society Duration: Dec 01, 2021 to Till Date Bangladesh Army University of Engineering and Technology (BAUET), Qadirabad Cantonment, Natore-6431 Duration: January 15, 2021 to till date
House Tutor, Bonolota Hall
Bangladesh Army University of Engineering and Technology (BAUET), Qadirabad Cantonment, Natore-6431 Duration: October, 2019 to August, 2024 Bangladesh Army University of Engineering and Technology (BAUET), Qadirabad Cantonment, Natore-6431 Duration: October, 2021 to till date
Former Assistant Proctor
Bangladesh Army University of Engineering and Technology (BAUET), Qadirabad Cantonment, Natore-6431
Former Course Coordinator
Bangladesh Army University of Engineering and Technology (BAUET), Qadirabad Cantonment, Natore-6431
Contact
Send a message to Bristi Rani Roy through the department office. Direct email and phone details are listed above.