efficientnetb2

oral_cancer_detection

COVID-19 CT scan image classification using EfficientNetB2 with transfer learning and deployment using Streamlit. This project focuses on accurately classifying CT scan images into three categories: COVID-19, Healthy, and Others. Leveraging transfer learning on pretrained EfficientNetB2 models, the classification model achieves robust performance.

Music emotions and themes classifier app could recognize 56 classes using three trained models (based on ResNet50, InceptionNetV2, EfficientNetB3), applying the transfer learning approach.

Deepfake detection using a hybrid CNN approach: spatial features from EfficientNet-B2, frequency analysis via FFT and ResNet-18, fused with cross-attention to reliably identify manipulated videos.

This repository presents a collaborative project on recycling material classification using a Convolutional Neural Network (CNN) based on EfficientNet-B2. It focuses on building an image classification system capable of identifying different types of recyclable materials, combining practical implementation with modern deep learning techniques.

This project uses the VGG16 and EfficientNetB2 models to recognize facial emotions in the FER-2013 dataset, classifying expressions into seven emotions.

Federated Alzheimer’s MRI classification on OASIS and ADNI with uniform, ROI-guided, adaptive, and layer-aware gradient perturbation, MIA evaluation, and PyTorch.

Implementation of Deep Learning Techniques in endoscopic images using Deep Learning