Publications

2024

Video-based Driver Distraction Detection using Transfer Learning: Bangladesh Perspective
Authors: Md. Muhaiminul Kabir, Md Abrar Nawar, Mehedi Islam Remon, Nishi Kanta Paul, Hanif Bhuiyan, Md Moinul Hoque
Conference: International Conference on Advances in Computing, Communication, Electrical, and Smart Systems (iCACCESS)
Publisher: IEEE Xplore
Abstract | PDF

Under Review — Conference Papers

  • [C2] N. K. Paul et al., "DR-LiteNet: A Lightweight Explainable Hybrid CNN Framework for Imbalanced Diabetic Retinopathy Grading via Adaptive SMOTE Fusion," in Proc. 1st Int. Conf. on Next-Generation Electrical & Electronics, Computer Systems, and Technologies (iCONEECT), 2026.
  • [C3] Shihabul, N. K. Paul et al., "BioVerify: Evidence-Aware Biomedical Answer Verification with Confidence-Based Uncertainty Detection for Patient-Safe Healthcare NLP," in Proc. iCONEECT, 2026.
  • [C4] N. K. Paul et al., "SleepEffFormer: Efficient CNN-Transformer with Transition-Aware Smoothing for Single-Channel EEG Sleep Stage Classification," in Proc. 2026 IEEE Int. Conf. on Biomedical Engineering, Computer and Information Technology for Health (BECITH), 2026.
  • [C5] Shihabul, N. K. Paul et al., "WaveFoG: Wavelet-Gated Transformer for Parkinson's Freezing of Gait Detection Using Wearable Accelerometer Signals," in Proc. IEEE BECITH, 2026.

Ongoing Research Projects

  • Radiomics-Conditioned Dual-Level Attention Fusion for Skin Lesion Classification — Target: Biomedical Signal Processing and Control. First Author · Elite Research Lab Collaboration. Developing a multi-stream dermoscopic classification framework using EfficientNet-B0, Swin Transformer-T, handcrafted radiomics, and DermTexNet with a dual-level radiomics-conditioned attention fusion module (DL-RCAF). Validation on ISIC 2019, HAM10000, PH2, and Derm7pt.
  • BGD-SF PolySegNet: Boundary-Guided Polyp Segmentation in Colonoscopy Images — Target: Journal Publication. First Author · Elite Research Lab Collaboration. Developing a boundary-guided dynamic selective fusion network for robust polyp segmentation with multi-scale feature aggregation, XAI-based interpretability, and calibration analysis.
  • Anomaly-Aware ForensiBlock — Washington State University Collaboration, supervised by Prof. Asma Jodeiri Akbarfam. Contributing to AI/ML and XAI pipeline for real-time authorized-user anomaly detection with asynchronous XAI-integrated authentication and blockchain-based audit logging.
  • Explainable Early-Stage Alzheimer's Disease Prediction — Supervising a university-level research project on early-stage Alzheimer's prediction using longitudinal clinical and MRI-derived features with interpretable machine learning techniques.