Search

{G}{ustav} Fechner's 1860 delineation of psychophysics, the measurement of sensation in relation to its stimulus, is widely considered to be the advent of modern psychological science. In psychophysics, a researcher parametrically varies some aspects of a stimulus, and measures the resulting changes in a human subject's experience of that stimulus; doing so gives insight to the determining relationship between a sensation and the physical input that evoked it. This approach is used heavily in perceptual domains, including signal detection, threshold measurement, and ideal observer analysis. Scientific fields like vision science have always leaned heavily on the methods and procedures of psychophysics, but there is now growing appreciation of them by machine learning researchers, sparked by widening overlap between biological and artificial perception \cite{rojas2011automatic, scheirer2014perceptual,escalera2014chalearn,zhang2018agil, grieggs2021measuring}. Machine perception that is guided by behavioral measurements, as opposed to guidance restricted to arbitrarily assigned human labels, has significant potential to fuel further progress in artificial intelligence., Comment: 6 pages, 3 figures, 1 table

Being able to distinguish between true DNA variants and technical sequencing artefacts is a fundamental task in whole genome, exome or targeted gene analysis. Variant calling tools provide diagnostic parameters, such as strand bias or an aggregated overall quality for each called variant, to help users make an informed choice about which variants to accept or discard. Having several such quality indicators poses a problem for the users of variant callers because they need to set or adjust thresholds for each such indicator. Alternatively, machine learning methods can be used to train a classifier based on these indicators. This approach needs large sets of labeled training data, which is not easily available. The new approach presented here relies on the idea that a true DNA variant exists independently of technical features of the read in which it appears (e.g. base quality, strand, position in the read). Therefore the nucleotide separability classification problem - predicting the nucleotide state of each read in a given pileup based on technical features only - should be near impossible to solve for true variants. Nucleotide separability, i.e. achievable classification accuracy, can either be used to distinguish between true variants and technical artefacts directly, using a thresholding approach, or it can be used as a meta-feature to train a separability-based classifier. This article explores both possibilities with promising results, showing accuracies around 90%.

Being able to distinguish between true DNA variants and technical sequencing artefacts is a fundamental task in whole genome, exome or targeted gene analysis. Variant calling tools provide diagnostic parameters, such as strand bias or an aggregated overall quality for each called variant, to help users make an informed choice about which variants to accept or discard. Having several such quality indicators poses a problem for the users of variant callers because they need to set or adjust thresholds for each such indicator. Alternatively, machine learning methods can be used to train a classifier based on these indicators. This approach needs large sets of labeled training data, which is not easily available. The new approach presented here relies on the idea that a true DNA variant exists independently of technical features of the read in which it appears (e.g. base quality, strand, position in the read). Therefore the nucleotide separability classification problem - predicting the nucleotide state of each read in a given pileup based on technical features only - should be near impossible to solve for true variants. Nucleotide separability, i.e. achievable classification accuracy, can either be used to distinguish between true variants and technical artefacts directly, using a thresholding approach, or it can be used as a meta-feature to train a separability-based classifier. This article explores both possibilities with promising results, showing accuracies around 90%.

Classical convolutional neural networks (cCNNs) are very good at categorizing objects in images. But, unlike human vision which is relatively robust to noise in images, the performance of cCNNs declines quickly as image quality worsens. Here we propose to use recurrent connections within the convolutional layers to make networks robust against pixel noise such as could arise from imaging at low light levels, and thereby significantly increase their performance when tested with simulated noisy video sequences. We show that cCNNs classify images with high signal to noise ratios (SNRs) well, but are easily outperformed when tested with low SNR images (high noise levels) by convolutional neural networks that have recurrency added to convolutional layers, henceforth referred to as gruCNNs. Addition of Bayes-optimal temporal integration to allow the cCNN to integrate multiple image frames still does not match gruCNN performance. Additionally, we show that at low SNRs, the probabilities predicted by the gruCNN (after calibration) have higher confidence than those predicted by the cCNN. We propose to consider recurrent connections in the early stages of neural networks as a solution to computer vision under imperfect lighting conditions and noisy environments; challenges faced during real-time video streams of autonomous driving at night, during rain or snow, and other non-ideal situations., Comment: 12 pages (with appendix), 6 figure (main text), 3 supplementary figures