O'TRAIN: a robust and flexible Real/Bogus classifier for the study of the optical transient sky

Abstract

Context. The scientific interest in studying high-energy transient phenomena in the Universe has largely grown for the last decade. Now, multiple ground-based survey projects have emerged to continuously monitor the optical (and multi-messenger) transient sky at higher image cadences and cover always larger portions of the sky every night. These novel approaches lead to a huge increase of the global alert rates which need to be handled with care especially by keeping the false alarms as low as possible. Therefore, the standard transient detection pipelines previously designed for narrow field of view instruments must now integrate more sophisticated tools to deal with the growing number and diversity of alerts and false alarms. Aims. Deep machine learning algorithms have now proven their efficiency in recognizing patterns in images. These methods are now used in astrophysics to perform different classification tasks such as identifying bogus from real transient point like sources. We explore this method to provide a robust and flexible algorithm that could be included in any kind of transient detection pipeline. Methods. We built a Convolutional Neural Network (CNN) algorithm in order to perform a real/bogus classification task on transient candidate cutouts (subtraction residuals) provided by different kinds of optical telescopes. The training involved human-supervised labeling of the cutouts, which had been split in two balanced data sets with True and False point-like source candidates. We tested our CNN model on the candidates produced by two different transient detection pipelines. In addition, we provide several diagnostic tools to evaluate the classification performance of our CNN models. Results. We show that our CNN algorithm can be successfully trained on a large diversity of images having very different pixel scales. In this training process, we did not detect any strong over(under)fitting with the requirement of providing cutouts with a limited size no larger than 50 × 50 pixels. Tested on optical images from four different telescopes and utilising two different transient detection pipelines, our CNN model provides robust real/bogus classification performance accuracy from 93% up to 98% of well classified candidates.

The effects of topological features on convolutional neural networks—an explanatory analysis via Grad-CAM

“…This problem is particularly important for the identification of the electromagnetic (EM) counterpart of a Table 1. The architecture of O'TRAIN [36] (top) and WaveNet [37] (bottom) before the classification head. (i: the layer number, k: the filter size, nc i : the number of filters, p: the dropout rate, r dilated : the rate of dilation).…”

“…We conducted Grad-CAM analysis on the CNN-BS Net. The architecture of the image network follows that of O'TRAIN [36]. O'TRAIN performs well on various grayscale images, particularly on the transient versus bogus image classification problem, and is widely used in the astrophysical image analysis community.…”