This project addresses the binary classification of dermoscopic skin lesion images as benign or malignant using deep learning techniques. Two different models are explored: a baseline convolutional neural network (CNN) and a ResNet50-based model using transfer learning.
The dataset used in this project is publicly available on Kaggle:
Fanconic. Skin Cancer: Malignant vs Benign
https://www.kaggle.com/datasets/fanconic/skin-cancer-malignant-vs-benign
Due to its size, the dataset is not included in this repository. Instead, it is automatically downloaded within the notebook using the gdown library from a public Google Drive link. Once downloaded, the dataset is extracted and organized into training and test folders.
notebooks/: Colab notebook with full pipeline (data loading, training, evaluation)requirements.txt: This project was developed and executed using Google Colab. The required libraries are listed here.figures/: Plots used in the presentation.report: The project report.presentation: Slides used in the presentation.