DeepExposureNet

# DeepExposure: A Multimodal Framework for Environmental Exposure Risk Prediction [](https://opensource.org/licenses/MIT)[](https://www.python.org/) ## 🧠 Project Overview **DeepExposure** is a multimodal deep learning framework designed to model and predict individual health risks in response to environmental exposure. Leveraging structured covariates, spatiotemporal features, and neural attention mechanisms, this project provides a powerful and interpretable pipeline for assessing environmental impact on respiratory outcomes. This repository accompanies our [Scientific Reports](https://www.nature.com/srep/) submission and includes all source code, dataset loaders, and pre-trained models for reproducibility and further development. ## 📂 Repository Structure