Project Overview: The project focuses on classifying brain tumors using the Multi-Modal Squeeze and Excitation Network.
Modalities and Pretrained Models: Two modalities are employed, utilizing pre-trained models: InceptionV2 and Xception net.
Hardware Requirements: The project requires GPU acceleration, and the Kaggle accelerator was utilized for its computational capabilities.
Dataset Source: The dataset is sourced from the public repository Figshare dataset.
Execution Environment: To run the code effectively, it's recommended to use the Kaggle platform, particularly with GPU support, as the model size exceeds the capacity of local CPUs.
Accessing the Dataset: The dataset is available on Kaggle, allowing users to easily access and run the code directly on the Kaggle platform.