Pediatric Pneumonia diagnosis using stacked ensemble learning on architectures on multi-modal deep CNN architectures.

The work proposes Contrast Limited Adaptive Histogram Equalization for image enhancement and a stacking classifier based on the fusion of deep learning-based features for pediatric pneumonia diagnosis. 

 Published in Multimedia tools and Application with Scopus Index Q1(81%) Journal Publication.

Title : Transfer Learning Approach for pediatric pneumonia diagnosis using channel attention deep CNN architectures.

This work proposes a modified attention mechanism that is inspired by Squeeze and Excitation architecture and stacking ensemble learning on features extracted from deep CNN channel attention architectures. 

Journal Publised in Engineering Applications of Artificial Intelligence.

Title : Pediatric Pneumonia diagnosis using stacked ensemble learning on architectures on multi-modal deep CNN architectures.

The work proposes a Convolutional Block Attention Module (CBAM) attached to the end of pre-trained ResNet152V2 and VGG19 with cost-sensitive learning.

Conference(accepted and presented), will be published soon. (link will be attached after publishing) 

Weighted Average Ensemble Approach for Pediatric Pneumonia Diagnosis using Channel Attention Deep CNN Architectures. 

Our proposed model incorporates a Channel Attention (ECA) module into pre-trained ResNet50, DenseNet121, and VGG19 models. We further enhance the performance by using a weighted average ensemble. The model achieved impressive results, including an accuracy of 95.67%, precision of 94.81%, recall of 98.46%, F1 score of 96.60%, and an AUC curve of 94.74% on the pediatric pneumonia dataset.

Conference(accepted), will be published soon. (link will be attached after publishing) .