Publications

13 peer-reviewed and under-review works. First-author papers are marked. Full author list as printed; Md Imam Ahasan bolded throughout.

Computer Vision
Medical Imaging
Diffusion Models
Generative Models
Scientific Machine Learning

2027

[R1]First AuthorCCF-BUnder Review

Dose and Anatomy-Aware Diffusion for Low-Dose CT Reconstruction

Md Imam Ahasan, Chengliang Wang, A F M Abdun Noor, Guangchao Yang

IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP 2027)

Extends low-dose CT reconstruction to generalize across unseen radiation doses and body regions without retraining.

PDF Code

[R8]CCF-BUnder Review

DeltaPath: Gated Delta-Rule Encoding for Efficient Vehicle Trajectory Representation Learning

Md Ariful Islam, Ruiyuan Li, A F M Abdun Noor, Md Imam Ahasan

IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP 2027)

Learns compact, efficient representations of vehicle trajectories for downstream mobility analysis.

2026

[C1]First AuthorCCF-CAccepted

Dose-Aware Cold Diffusion with Physics Consistency for Generalizable Low-Dose CT Reconstruction

Md Imam Ahasan, Guangchao Yang, A F M Abdun Noor, S M Hasan Mahmud, Md Mahfuzur Rahman

Intl. Joint Conf. on Neural Networks (IJCNN 2026)

Reconstructs clean CT images from noisy low-dose scans more accurately and 3.3× faster than prior methods.

PDF Code

[C3]EIAccepted

MTF-Net: Multi-Modal Temporal Feature Fusion Network for Pedestrian Intention Prediction

Md Mahfuzur Rahman, Pengzhan Zhou, A F M Abdun Noor, Md Imam Ahasan

Intl. Conf. on Cloud and Network Computing (ICCNC 2026)

Predicts pedestrian crossing intent from video and sensor data to help autonomous vehicles react earlier.

PDF Code

[C4]EIAccepted

MoSSGate: Memory-Modulated State-Space Gating for Skin Lesion Segmentation

Anum Malik, Mahnoor Buriro, Muhammad Younas Khan, Md Imam Ahasan

Intl. Conf. on Cloud and Network Computing (ICCNC 2026)

Improves automatic detection of skin lesion boundaries in dermatology images.

PDF Code

[J1]SCIEAccepted

Direction-Curvature Aware Feature Integration for Robust Lane Detection

Ahtisham Waheed, Yunfei Yin, A F M Abdun Noor, Md Imam Ahasan

Computers, Materials and Continua (CMC)

Detects road lane markings more robustly on curved and poorly marked roads for autonomous driving.

DOI Code

[R2]CCF-CUnder Review

TSCA-Net: Temporal-Spatial Clique Attention for Interpretable Multimodal Pedestrian Trajectory Prediction

Md Mustafizur Rahman, Guangchao Yang, A F M Abdun Noor, Md Imam Ahasan

Intl. Conf. on Mobility, Sensing and Networking (MSN 2026)

Forecasts where pedestrians will walk next from multiple sensor views, with interpretable attention maps.

PDF Code

[R3]First AuthorSCIEUnder Review

LightGAN-LD: A Lightweight Generative Adversarial Network for Efficient Low-Dose CT Reconstruction with Sinogram Encoding and Edge-Aware Learning

Md Imam Ahasan, Guangchao Yang, A F M Abdun Noor, Mohammad Azam Khan

PeerJ Computer Science

A lighter, faster GAN for low-dose CT reconstruction suited to real-time clinical deployment.

Code

[J2]First AuthorSCIEAccepted

Toward Real-Time LDCT Reconstruction: A Sinogram-Encoded Lightweight GAN with Measurement-Domain Noise Suppression

Md Imam Ahasan, A F M Abdun Noor, Kah Ong Michael Goh, S M Hasan Mahmud

Computers, Materials and Continua (CMC)

Suppresses noise directly in the raw scanner data for real-time low-dose CT reconstruction.

[J3]SCIEAccepted

BiLoG-Net: A Bi-Context Location-Guided Network for Breast Mass Segmentation and Malignancy Classification in Mammography

A F M Abdun Noor, Md Imam Ahasan, Kah Ong Michael Goh, S M Hasan Mahmud

Computer Modeling in Engineering and Sciences (CMES)

Jointly locates and classifies findings in mammograms to support breast cancer screening.

PDF Code

[R6]SCIEUnder Review

Adaptive Cross-Modal Fusion with Sparse Attention for Pedestrian Crossing Intention Prediction

Md Mahfuzur Rahman, Pengzhan Zhou, A F M Abdun Noor, Md Imam Ahasan

PeerJ Computer Science

Combines multiple sensor modalities to predict pedestrian crossing intent more robustly.

PDF Code

[R7]First AuthorSCIEUnder Review

PCDiff: Physics-Conditioned Bayesian Diffusion Model for Low-Dose CT Reconstruction

Md Imam Ahasan, Guangchao Yang, A F M Abdun Noor, Mohammad Azam Khan

Computer Modeling in Engineering and Sciences (CMES)

A Bayesian diffusion model that conditions reconstruction on the CT scanner's physics to better quantify uncertainty in low-dose CT images.

Code

2025

[C2]CCF-CPublished

GeGLUNet: Structural Retinal Vessel Segmentation via Attention-Gated GeGLU and Contrastive Supervision

A F M Abdun Noor, Md Imam Ahasan, Mohammad Azam Khan, Guangchao Yang

Conf. on Pattern Recognition and Computer Vision (PRCV 2025)

Segments blood vessels in retinal scans more precisely to support earlier diagnosis of eye disease.

DOI Code