9 papers
Region-aware error triage for medical image segmentation with groupwise conformal risk control
Nishi Kanta Paul, Md Shihabul Islam Shovo, Hoileong Lee
Segmentation masks are increasingly reviewed rather than redrawn, which requires knowing where a mask is likely wrong. Learned error maps rank pixels by risk, but the number of pix…
LCM-UNet: Scale-Selective Mamba with Reparameterizable Multi-Basis Local Modeling for Lightweight Skin Lesion Segmentation
Md Shihabul Islam Shovo, Nishi Kanta Paul, Dip Kumar Saha +1
Accurate lesion segmentation underpins automated dermoscopic analysis: masks enable measurements of asymmetry, border and size. Deployment in primary care or on mobile devices requ…
DR-LiteNet: A Lightweight Explainable Hybrid CNN Framework for Imbalanced Diabetic Retinopathy Grading via Adaptive SMOTE Fusion
Nishi Kanta Paul, Md Shihabul Islam Shovo, Audrija Chowdhury +1
Automated diabetic retinopathy (DR) grading from colour fundus photographs is held back by two compounding challenges: severe class imbalance across five severity grades and the he…
Band2Net: Frequency-to-Network Transport Distillation for fMRI-Privileged, EEG-Only Session-Condition Decoding
Nishi Kanta Paul, Md Shihabul Islam Shovo, Dwarikanath Mahapatra
Functional MRI resolves large-scale brain networks but is expensive and stationary, whereas EEG is portable but poorly localised. We treat fMRI as privileged information, available…
Coupled or Independent? Fault Coupling in Agentic Medical Vision Systems
Nishi Kanta Paul, Md Shihabul Islam Shovo, Md Mostafijur Rahman +1
Vision-language agents equipped with tools expose intermediate outputs for inspection and comparison with the model's reasoning, which may offer safety benefits over single-shot mo…
AquaProbe: Efficient Adaptation of Frozen Vision Models for Underwater Images
Md Shihabul Islam Shovo, Nishi Kanta Paul, Trung Dong +1
Reliable visual perception is essential for safe autonomous underwater exploration. Yet vision models pretrained on terrestrial images become less reliable underwater, where turbid…
LAPIS: Clinically Constrained Patient-Identity Residual Learning for Chest X-Ray Anonymization
Nishi Kanta Paul, Md Shihabul Islam Shovo, Désiré Sidibé
Biometric information in chest radiographs limits their release for research, as patients can be re-identified. We propose LAPIS, which jointly adapts a spatial chest-radiograph la…
SleepEffFormer: Efficient CNN-Transformer with Transition-Aware Smoothing for Single-Channel EEG Sleep Stage Classification
Nishi Kanta Paul, Md Shihabul Islam Shovo, Israt Jerin Esha +1
Automated sleep stage classification based on single-channel EEG is a promising pathway to large-scale sleep monitoring outside the polysomnographic lab. This paper proposes SleepE…
WaveFoG: Wavelet Gated Transformer for Parkinson's Freezing of Gait Detection Using Wearable Accelerometer Signals
Md Shihabul Islam Shovo, Nishi Kanta Paul, Adrita Rahman +1
Freezing of gait (FoG) is one of the most debilitating episodic motor symptoms of Parkinson's disease (PD). FoG detection from wrist worn inertial signals poses challenging problem…