Recognition of Cosmic Ray Images Obtained from CMOS Sensors Used in Mobile Phones by Approximation of Uncertain Class Assignment with Deep Convolutional Neural Network

Abstract

In this paper, we describe the convolutional neural network (CNN)-based approach to the problems of categorization and artefact reduction of cosmic ray images obtained from CMOS sensors used in mobile phones. As artefacts, we understand all images that cannot be attributed to particles’ passage through sensor but rather result from the deficiencies of the registration procedure. The proposed deep neural network is composed of a pretrained CNN and neural-network-based approximator, which models the uncertainty of image class assignment. The network was trained using a transfer learning approach with a mean squared error loss function. We evaluated our approach on a data set containing 2350 images labelled by five judges. The most accurate results were obtained using the VGG16 CNN architecture; the recognition rate (RR) was 85.79% ± 2.24% with a mean squared error (MSE) of 0.03 ± 0.00. After applying the proposed threshold scheme to eliminate less probable class assignments, we obtained a RR of 96.95% ± 1.38% for a threshold of 0.9, which left about 62.60% ± 2.88% of the overall data. Importantly, the research and results presented in this paper are part of the pioneering field of the application of citizen science in the recognition of cosmic rays and, to the best of our knowledge, this analysis is performed on the largest freely available cosmic ray hit dataset.

CNN-Based Classifier as an Offline Trigger for the CREDO Experiment

“…The possibilities of using the concept of transfer learning were also analysed, where such networks are pretrained for large, standardised data sets, such as ImageNet [ 48 ]. The transfer learning approach to classifying the CREDO data was already discussed in [ 49 ]. Due to the peculiarity of the problem, quite unusual input data and a small spatial size of the signal in the images (only a few to a maximum of several dozen pixels), we decided to develop a dedicated architecture tailored to the specifics and requirements of the problem.…”

Zernike Moment Based Classification of Cosmic Ray Candidate Hits from CMOS Sensors

“…Recent analyses of data acquired by the CREDO experiment employed various CNN architectures to detect potentially relevant signals [12] and classify them [32]. Here, rather than CNN based classifiers, we discuss an approach based on the classical statistical learning classifiers implemented in the sci-kit-learn library [31].…”

“…In the present work, the MLP classifiers with one-vs-rest multiclass strategy provide good recognition capabilities for all four classes of events with small contamination from wrong identification. These results are compared in Table 7 with results from [13,32]. There is a certain improvement in the Track recognition capability with the new approach when compared to [32] without too much deterioration in identifying the Worms.…”

“…These results are compared in Table 7 with results from [13,32]. There is a certain improvement in the Track recognition capability with the new approach when compared to [32] without too much deterioration in identifying the Worms. Although, the contamination of worms identified as tracks and vice-versa is reduced.…”

SLAM-OR: Simultaneous Localization, Mapping and Object Recognition Using Video Sensors Data in Open Environments from the Sparse Points Cloud

“…Application of DNN has revolutionized image-based procedures of features generation, objects detection and recognition [ 26 , 27 ]. This brain-inspired machine learning algorithms [ 28 , 29 ] have found their application, i.e., for security purposes (face detection, vehicle detection) [ 30 , 31 , 32 ], they are used in autonomous cars to recognize the environment [ 33 ], in medicine [ 34 ] etc.…”