Enhanced Rotational Invariant Convolutional Neural Network for Supernovae Detection

TL;DR

This paper introduces an enhanced CNN with rotational invariance and visualization techniques for supernovae detection, achieving high accuracy and better interpretability compared to previous models.

Contribution

The paper presents a new rotational invariance method and an LRP-based visualization approach, improving supernova detection accuracy and interpretability.

Findings

Achieved 99.53% accuracy on HiTS dataset.

Enhanced rotational invariance compared to original model.

Improved interpretability through relevance heatmaps.

Abstract

In this paper, we propose an enhanced CNN model for detecting supernovae (SNe). This is done by applying a new method for obtaining rotational invariance that exploits cyclic symmetry. In addition, we use a visualization approach, the layer-wise relevance propagation (LRP) method, which allows finding the relevant pixels in each image that contribute to discriminate between SN candidates and artifacts. We introduce a measure to assess quantitatively the effect of the rotational invariant methods on the LRP relevance heatmaps. This allows comparing the proposed method, CAP, with the original Deep-HiTS model. The results show that the enhanced method presents an augmented capacity for achieving rotational invariance with respect to the original model. An ensemble of CAP models obtained the best results so far on the HiTS dataset, reaching an average accuracy of 99.53%. The improvement over Deep-HiTS is significant both statistically and in practice.