A web application that classifies an uploaded image into a disaster type, utilizing Angular

Deep Learning for SAR Ship classification: Focus on Unbalanced Datasets and Inter-Dataset Generalization

Histopathologic metastatic breast cancer detection with convolution neural networks on pathology whole slide images using TensorFlow.

One of the firsts dataset level explanability libraries for 1d signal using GRAD-CAM++

Multi-classification of ocular diseases using ODIR-5K and uses explainable AI methods to improve the transparency of computer aided diagnostic systems.

Fellowship.AI Challenge: Cohort 19

Visualizing the decision making process of models - for melanoma detection

Implementation of Gradient-weighted Class Activation Maps of a simple classifier in Tensorflow. In progress.

Medical Diagnosis in Chest X-ray images.

This notebook demonstrates a number of saliency mask techniques, augmented with the SmoothGrad technique, using the VGG16, AlexNet, InceptioNet, MobileNet convolutional neural network in TF2.

Diagnose the presence of skin cancer in a person using CNN and as well explain what led the CNN to arrive at the decision. Visual explanations are made utilizing the Gradient-weighted Class Activation Mapping (Grad-CAM), the gradients flowing into the final convolutional layer to produce a coarse localization map highlighting the important regio…

In this repo. I apply several variation of GradCAM including GradCAM, GradCAMPlusPlus, EigenCAM and etc on a pretrained Resnet model and report ROAD as an evaluation metric for interpretability.

Grad-CAM implementation in TensorFlow JS.

This repository contains a CIFAR-10 dataset classification project alongside GradCAM implementation

Analyzing the Effect of Adversarial Inputs on Saliency Maps

Class Activation Map (CAM and Grad-CAM) Analysis of fine-tuned CNNs with transfer learning for Pokemon classification task to understand the features learned by deep CNN

ISIC2019 skin lesion classification (binary & multi-class) as well as segmentation pipelines using VGG16_BN and visual attention blocks. The project features improving the results found in the literature by implementing an ensemble architecture. This project was developed for "Computer Aided Diagnosis - CAD" course for MAIA masters program.

Analysis of visualization methods for relevant areas within images for a trained CNN. Used the GTSRB dataset as well as Activation Maximization, Saliency Map, GradCam, and Gradcam++ methods.

Chest x-ray classifier to diagnose between covid-19 infected lungs, viral pneumonia and healthy lungs using deep learning.