Project Overview

Skin cancer is a crucial health issue that requires timely detection for higher survival rates. Traditional computer vision techniques face challenges in addressing the advanced variability of skin lesion features, a gap partially bridged by convolutional neural networks (CNNs). To overcome the existing issues, we introduce an innovative convolutional ensemble network approach named deep autoencoder (DAE) with VGG19 & ResNet101. This method utilizes convolution-based deep neural networks for the detection of skin cancer. The ISIC-2018 public data taken from the source is used for experimental results, which demonstrate remarkable performance with the different in terms of performance metrics.

Model Architecture

• 0: akiec: Actinic keratoses and intraepithelial carcinoma
• 1: bcc: Basal cell carcinoma
• 2: bkl: Benign keratosis-like lesions
• 3: df: Dermatofibroma
• 4: mel: Melanoma
• 5: nv: Melanocytic nevi
• 6: vasc: Vascular lesions

Class	Example
akiec
bcc
bkl
df
mel
nv
vasc

Play with Skin Cancer Images(https://www.kaggle.com/datasets/kmader/skin-cancer-mnist-ham10000)

    📂 Dataset

The dataset used in this project is the HAM10000 dataset.

• Source: HAM10000 Dataset
• Images: 10,015 dermoscopic images, categorized into seven classes.
• Preprocessing: Images resized to 128x128 for consistency.

---

Overview

• Goal: Predicted classification of the skin cancer of 7 class.
• Dataset: Skin cancer datasets with 7 class.

---

Instructions

1. Run the Code:

1. Run `SKIN_ResNet50_VGG19.ipynb'.

2. Dataset:

• Input file: HAM10000 dataset.

---

Results

• Accuracy, F1 Score, AUC(ROC)

---

Libraries Used:

• numpy
• keras
• tensorflow-cpu==2.5.0
• pandas
• matplotlib
• pillow
• flask
• seaborn
• gunicorn

---

Model Summary

Acknowledgements

• UpGrad tutorials on Convolution Neural Networks (CNNs) on the learning platform
• Melanoma Skin Cancer
• Introduction to CNN
• Image classification using CNN
• Efficient way to build CNN architecture

---