Deep Learning Architectures: ResNet, Xception, and DenseNet

Project Overview

This repository explores 3 deep learning architectures for classification tasks:

1. ResNet: Implemented from scratch.
2. Xception: Transfer learning with fine-tuning.
3. DenseNet: Transfer learning with fine-tuning.

---

Architectures Overview

ResNet (Residual Networks)

• Introduced in "Deep Residual Learning for Image Classification" by He et al. (2015).
• Key Idea: Skip connections to combat vanishing gradients.
• Implemented from scratch in this project.

Xception

• Presented in "Xception: Deep Learning with Depthwise Separable Convolutions" by Chollet (2017).
• Focuses on depthwise separable convolutions to reduce computation.

DenseNet (Densely Connected Networks)

• Proposed in "Densely Connected Convolutional Networks" by Huang et al. (2017).
• Ensures feature reuse by connecting each layer to all others.

---

Evaluation Metrics

1. Accuracy: Measures the percentage of correctly classified samples.
2. Confusion Matrix: Visualized to analyze true/false positives and negatives.
3. Recall, Precision, F1-Score: Highlights the model's performance balance.
4. ROC and AUC: Visualizes the trade-off between sensitivity and specificity.

---

References (references updated and the actual references for the project are mentioned in the documentation file)

1. ResNet: He et al., 2015 Deep Residual Learning for Image Recognition
2. Xception: Chollet, 2017 Xception: Deep Learning with Depthwise Separable Convolutions
3. DenseNet: Huang et al., 2017 Densely Connected Convolutional Networks