Rapid and reliable identification of insects is important in many contexts, from the detection of disease vectors and invasive species to the sorting of material from biodiversity inventories. Because of the shortage of adequate expertise, there has long been an interest in developing automated systems for this task. Previous attempts have been based on laborious and complex handcrafted extraction of image features, but in recent years it has been shown that sophisticated convolutional neural networks (CNNs) can learn to extract relevant features automatically, without human intervention. Unfortunately, reaching expert-level accuracy in CNN identifications requires substantial computational power and huge training data sets, which are often not available for taxonomic tasks. This can be addressed using feature transfer: a CNN that has been pretrained on a generic image classification task is exposed to the taxonomic images of interest, and information about its perception of those images is used in training a simpler, dedicated identification system. Here, we develop an effective method of CNN feature transfer, which achieves expert-level accuracy in taxonomic identification of insects with training sets of 100 images or less per category, depending on the nature of data set. Specifically, we extract rich representations of intermediate to high-level image features from the CNN architecture VGG16 pretrained on the ImageNet data set. This information is submitted to a linear support vector machine classifier, which is trained on the target problem. We tested the performance of our approach on two types of challenging taxonomic tasks: 1) identifying insects to higher groups when they are likely to belong to subgroups that have not been seen previously and 2) identifying visually similar species that are difficult to separate even for experts. For the first task, our approach reached \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$CDATA[$CDATA[$>$$\end{document}92% accuracy on one data set (884 face images of 11 families of Diptera, all specimens representing unique species), and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$CDATA[$CDATA[$>$$\end{document}96% accuracy on another (2936 dorsal habitus images of 14 families of Coleoptera, over 90% of specimens belonging to unique species). For the second task, our approach outperformed a leading taxonomic expert on one data set (339 images of three species of the Coleoptera genus Oxythyrea; 97% accuracy), and both humans and traditional automated identification systems on another data set (3845 images of nine species of Plecoptera larvae; 98.6 % accuracy). Reanalyzing several biological imag...

No abstract

We report a new test of quantum electrodynamics (QED) for the w (1s2p 1 P1 → 1s 2 1 S0) X-ray resonance line transition energy in helium-like titanium. This measurement is one of few sensitive to two-electron QED contributions. Systematic errors such as Doppler shifts are minimised in our experiment by trapping and stripping Ti atoms in an Electron Beam Ion Trap (EBIT) and by applying absolute wavelength standards to calibrate the dispersion function of a curved-crystal spectrometer. We also report a more general systematic discrepancy between QED theory and experiment for the w transition energy in helium-like ions for Z > 20. When all of the data available in the literature forZ = 16 − 92 is taken into account, the divergence is seen to grow as approximately Z 3 with a statistical significance on the coefficient that rises to the level of five standard deviations. Our result for titanium alone, 4749.85(7) eV for the w-line, deviates from the most recent ab initio prediction by three times our experimental uncertainty and by more than ten times the currently estimated uncertainty in the theoretical prediction.PACS numbers: 31.30.jf, 12.20.Fv, 34.50.Fa, 32.30.Rj Quantum electrodynamics (QED) is a cornerstone of modern theoretical physics. New activity on this topic has been stimulated by the announcement of a five-sigma inconsistency between a 15 ppm (parts per million) measurement of an atomic transition frequency in muonic hydrogen [1] and independent measurements of the proton size, linked together by QED calculations. The high sensitivity of such a measurement to QED is derived in part from the large mass of the bound lepton which shrinks the orbital radius. Another way to reduce the orbital radius and study magnified QED effects is to measure transitions in highly charged ions of increasing Z. QED processes scale as various powers of Zα and significantly affect the quantum observable, namely transition energies. Moreover, in the high-Z range, some of the perturbative expansions fail, so that theoretical methods very different from those used for hydrogen are required. Since QED treatment of low-Z and high-Z systems are undertaken with significantly different starting points and mathematical techniques, precise measurements for ions in the mid-Z range will guide the long-pursued development of a unified computational methodology with very accurate predictions for the entire domain Z < 100 [2,3].Advances in QED theory have been sufficient that one can go beyond one-lepton systems (either free or bound) and explore the three-body quantum problem to high precision, including the investigation of helium-like * Electronic address: chantler@unimelb.edu.au atomic systems with two electrons bound to a nucleus. Here the two-electron QED contributions that are entirely absent in one-electron systems can be probed and compared to various theoretical formulations. In this work, we report a measurement of the strongest resonant transition 1s2p1 P 1 → 1s 2 1 S 0 in He-like Ti (Ti 20+ ), and present a divergence that is...

No abstract

No abstract