This project aims to detect Intrusions with a network using deep learning. The network traffic data is converted to multi channel RGB images, that are passed through CNNs to detect features useful to intrusion detection. (Additionally, we also experimented with dense SIFT based feature description. To discuss more on it, feel free to reach out.)