Institution profile

Bangladesh University of Business & Technology

Academic institutionasia · bd
Official website
Research library17linked papers
Opportunities0open roles
Selected work

Representative Papers

Privileged Lesion-Context Relational Distillation for Mask-Free Skin Lesion Classification

Jul 21, 2026

This work addresses the challenge of skin lesion classification, which often relies on segmentation masks or auxiliary models that hinder clinical deployment and increase computational overhead. The authors propose the PLCRD framework, which leverages lesion masks during training to construct a teacher model and transfers structured knowledge about lesion–context relationships to a student model that requires only raw images for inference, thereby enabling mask-free prediction. Innovatively, privileged mask information is transformed into transferable relational knowledge, circumventing direct feature alignment between heterogeneous architectures. The approach integrates multiple mechanisms—including diagnostic distribution transfer, attention propagation, lesion similarity alignment, and lesion–context affinity matching. Evaluated on HAM10000 and ISIC 2018, the method achieves macro F1 scores of 0.773 ± 0.018 and 0.732 ± 0.008, respectively, significantly advancing classification performance under mask-free conditions.

0 citationsRead paper

Simultaneous Detection of LSD and FMD in Cattle Using Ensemble Deep Learning

Jan 19, 2026

This study addresses the diagnostic challenge posed by the high clinical similarity between lumpy skin disease (LSD) and foot-and-mouth disease (FMD), which are often confused with benign dermatological conditions, leading to delayed detection and control. To tackle this issue, the authors propose a novel ensemble deep learning framework for simultaneous multi-disease classification, integrating VGG16, ResNet50, and InceptionV3 architectures through transfer learning and an optimized weighted averaging strategy. The model was trained and validated on a dataset of 10,516 expert-annotated images, achieving 98.2% accuracy, 98.1% macro-averaged recall and F1-score, and an AUC-ROC of 99.5% in concurrent LSD and FMD recognition. These results significantly outperform existing approaches, demonstrating enhanced accuracy and practical utility for early diagnosis in field settings.

0 citationsRead paper
Recent publications

Latest Papers

Privileged Lesion-Context Relational Distillation for Mask-Free Skin Lesion Classification

Jul 21, 2026

This work addresses the challenge of skin lesion classification, which often relies on segmentation masks or auxiliary models that hinder clinical deployment and increase computational overhead. The authors propose the PLCRD framework, which leverages lesion masks during training to construct a teacher model and transfers structured knowledge about lesion–context relationships to a student model that requires only raw images for inference, thereby enabling mask-free prediction. Innovatively, privileged mask information is transformed into transferable relational knowledge, circumventing direct feature alignment between heterogeneous architectures. The approach integrates multiple mechanisms—including diagnostic distribution transfer, attention propagation, lesion similarity alignment, and lesion–context affinity matching. Evaluated on HAM10000 and ISIC 2018, the method achieves macro F1 scores of 0.773 ± 0.018 and 0.732 ± 0.008, respectively, significantly advancing classification performance under mask-free conditions.

0 citationsRead paper

Simultaneous Detection of LSD and FMD in Cattle Using Ensemble Deep Learning

Jan 19, 2026

This study addresses the diagnostic challenge posed by the high clinical similarity between lumpy skin disease (LSD) and foot-and-mouth disease (FMD), which are often confused with benign dermatological conditions, leading to delayed detection and control. To tackle this issue, the authors propose a novel ensemble deep learning framework for simultaneous multi-disease classification, integrating VGG16, ResNet50, and InceptionV3 architectures through transfer learning and an optimized weighted averaging strategy. The model was trained and validated on a dataset of 10,516 expert-annotated images, achieving 98.2% accuracy, 98.1% macro-averaged recall and F1-score, and an AUC-ROC of 99.5% in concurrent LSD and FMD recognition. These results significantly outperform existing approaches, demonstrating enhanced accuracy and practical utility for early diagnosis in field settings.

0 citationsRead paper