Food Recognition for Dietary Assessment Using Deep Convolutional Neural Networks

Diet management is a key factor for the prevention and treatment of diet-related chronic diseases. Computer vision systems aim to provide automated food intake assessment using meal images. We propose a method for the recognition of already segmented food items in meal images. The method uses a 6-layer deep convolutional neural network to classify food image patches. For each food item, overlapping patches are extracted and classified and the class with the majority of votes is assigned to it. Experiments on a manually annotated dataset with 573 food items justified the choice of the involved components and proved the effectiveness of the proposed system yielding an overall accuracy of 84.9%.

Reproductive Biology of the Deep-Sea Polychaete Gorgoniapolynoe Caeciliae (Polynoidae), a Commensal Species Associated with Octocorals

Some aspects of the reproductive biology of the polychaete Gorgoniapolynoe caeciliae have been described for the first time. Gorgoniapolynoe caeciliae is a deep-sea commensal species associated with Candidella imbricala, all octocoral that populates the New England Seamount chain. Gorgoniapolynoe caeciliae is a dioccious species with an equal sex ratio and fertile segments throughout most of the adult body. The gonads of both sexes are associated with genital blood vessels emerging from the posterior surface of most intersegmental septa. In the female, oogenesis is intraovarian with oocytes being retained within the ovary until vitellogenesis is completed. The largest female examined contained over 3000 eggs with a maximum diameter of 80-90 mu m. In the male, the testes are repeated in numerous segments and consist of small clusters of spermatogonia, spermatocytes and early spermatids associated with the walls of the genital blood vessels. Early spermatids are shed into the coelom where they complete differentiation into mature ect-aquasperm with a spherical head (4 mu m), a small cap-like acrosome, and a short mid-piece with four mitochondria. Indirect evidence suggests that this species is an annual breeder that releases its gametes into seawater and produces a planktotrophic larva following fertilization. The reproductive biology of G. caeciliae is consistent with that of most other polynoids including many shallow water species suggesting that phylogenetic history strongly shapes its biology.

Static GPS measurements and elevation model in the vicinity of the EDML deep-drilling site

Interpretation of ice-core records requires accurate knowledge of the past and present surface topography and stress-strain fields. The European Project for Ice Coring in Antarctica (EPICA) drilling site (0.0684° E and 75.0025° S, 2891.7 m) in Dronning Maud Land, Antarctica, is located in the immediate vicinity of a transient and splitting ice divide. A digital elevation model is determined from the combination of kinematic GPS measurements with the GLAS12 data sets from the ICESat satellite. Based on a network of stakes, surveyed with static GPS, the velocity field around the EDML drilling site is calculated. The annual mean velocity magnitude of 12 survey points amounts to 0.74 m/a. Flow directions mainly vary according to their distance from the ice divide. Surface strain rates are determined from a pentagon-shaped stake network with one center point, close to the drilling site. The strain field is characterised by along flow compression, lateral dilatation, and vertical layer thinning.

Pore water geochemistry of arctic deep sea sediments measured at station MSM16/2_809-1 (REF, Reference)

Pore water and turnover rates were determined for surface sediment cores obtained in 2009 and 2010. The pore water was extracted with Rhizons (Rhizon CSS: length 5 cm, pore diameter 0.15 µm; Rhizosphere Research Products, Wageningen, Netherlands) in 1 cm-resolution and immediately fixed in 5% zinc acetate (ZnAc) solution for sulfate, and sulfide analyses. The samples were diluted, filtered and the concentrations measured with non-suppressed anion exchange chromatography (Waters IC-Pak anion exchange column, waters 430 conductivity detector). The total sulfide concentrations (H2S + HS- + S**2-) were determined using the diamine complexation method (doi:10.4319/lo.1969.14.3.0454). Samples for dissolved inorganic carbon (DIC) and alkalinity measurements were preserved by adding 2 µl saturated mercury chloride (HgCl2) solution and stored headspace-free in gas-tight glass vials. DIC and alkalinity were measured using the flow injection method (detector VWR scientific model 1054) (doi:10.4319/lo.1992.37.5.1113). Dissolved sulfide was eliminated prior to the DIC measurement by adding 0.5 M molybdate solution (doi:10.4319/lo.1995.40.5.1011). Nutrient subsamples (10 - 15 ml) were stored at - 20 °C prior to concentration measurements with a Skalar Continuous-Flow Analyzer (doi:10.1002/9783527613984).

Sulphate reduction rates of arctic deep sea sediments measured at station MSM16/2_809-1 (REF, Reference)

Turnover rates were determined for surface sediment cores obtained in 2009 and 2010. Sulfate reduction (SR) were measured ex situ by the whole core injection method (doi:10.1080/01490457809377722). We incubated the samples at in situ temperature (1.0°C) for 12 hours with carrier-free 35**SO4 (dissolved in water, 50 kBq). Sediment was fixed in 20 ml 20% ZnAc solution for AOM or SR, respectively. Turnover rates were measured as previously described (doi:10.4319/lom.2004.2.171).

Constraining storm-scale forecasts of deep convective initiation with surface weather observations

Thesis (Ph.D.)--University of Washington, 2016-06

Representation and learning of visual information for pose recognition

Recovering position from sensor information is an important problem in mobile robotics, known as localisation. Localisation requires a map or some other description of the environment to provide the robot with a context to interpret sensor data. The mobile robot system under discussion is using an artificial neural representation of position. Building a geometrical map of the environment with a single camera and artificial neural networks is difficult. Instead it would be simpler to learn position as a function of the visual input. Usually when learning images, an intermediate representation is employed. An appropriate starting point for biologically plausible image representation is the complex cells of the visual cortex, which have invariance properties that appear useful for localisation. The effectiveness for localisation of two different complex cell models are evaluated. Finally the ability of a simple neural network with single shot learning to recognise these representations and localise a robot is examined.

Deep brain stimulation for chronic pain investigated with magnetoencephalography

Deep brain stimulation has shown remarkable potential in alleviating otherwise treatment-resistant chronic pain, but little is currently known about the underlying neural mechanisms. Here for the first time, we used noninvasive neuroimaging by magnetoencephalography to map changes in neural activity induced by deep brain stimulation in a patient with severe phantom limb pain. When the stimulator was turned off, the patient reported significant increases in subjective pain. Corresponding significant changes in neural activity were found in a network including the mid-anterior orbitofrontal and subgenual cingulate cortices; these areas are known to be involved in pain relief. Hence, they could potentially serve as future surgical targets to relieve chronic pain. © 2007 Lippincott Williams & Wilkins, Inc.

Developing critical understanding in HRM students:using innovative teaching methods to encourage deep approaches to study

Purpose – This paper aims to focus on developing critical understanding in human resource management (HRM) students in Aston Business School, UK. The paper reveals that innovative teaching methods encourage deep approaches to study, an indicator of students reaching their own understanding of material and ideas. This improves student employability and satisfies employer need. Design/methodology/approach – Student response to two second year business modules, matched for high student approval rating, was collected through focus group discussion. One module was taught using EBL and the story method, whilst the other used traditional teaching methods. Transcripts were analysed and compared using the structure of the ASSIST measure. Findings – Critical understanding and transformative learning can be developed through the innovative teaching methods of enquiry-based learning (EBL) and the story method. Research limitations/implications – The limitation is that this is a single case study comparing and contrasting two business modules. The implication is that the study should be replicated and developed in different learning settings, so that there are multiple data sets to confirm the research finding. Practical implications – Future curriculum development, especially in terms of HE, still needs to encourage students and lecturers to understand more about the nature of knowledge and how to learn. The application of EBL and the story method is described in a module case study – “Strategy for Future Leaders”. Originality/value – This is a systematic and comparative study to improve understanding of how students and lecturers learn and of the context in which the learning takes place.

Background synaptic activity in rat entorhinal cortex shows a progressively greater dominance of inhibition over excitation from deep to superficial layers

The entorhinal cortex (EC) controls hippocampal input and output, playing major roles in memory and spatial navigation. Different layers of the EC subserve different functions and a number of studies have compared properties of neurones across layers. We have studied synaptic inhibition and excitation in EC neurones, and we have previously compared spontaneous synaptic release of glutamate and GABA using patch clamp recordings of synaptic currents in principal neurones of layers II (L2) and V (L5). Here, we add comparative studies in layer III (L3). Such studies essentially look at neuronal activity from a presynaptic viewpoint. To correlate this with the postsynaptic consequences of spontaneous transmitter release, we have determined global postsynaptic conductances mediated by the two transmitters, using a method to estimate conductances from membrane potential fluctuations. We have previously presented some of this data for L3 and now extend to L2 and L5. Inhibition dominates excitation in all layers but the ratio follows a clear rank order (highest to lowest) of L2>L3>L5. The variance of the background conductances was markedly higher for excitation and inhibition in L2 compared to L3 or L5. We also show that induction of synchronized network epileptiform activity by blockade of GABA inhibition reveals a relative reluctance of L2 to participate in such activity. This was associated with maintenance of a dominant background inhibition in L2, whereas in L3 and L5 the absolute level of inhibition fell below that of excitation, coincident with the appearance of synchronized discharges. Further experiments identified potential roles for competition for bicuculline by ambient GABA at the GABAA receptor, and strychnine-sensitive glycine receptors in residual inhibition in L2. We discuss our results in terms of control of excitability in neuronal subpopulations of EC neurones and what these may suggest for their functional roles. © 2014 Greenhill et al.

Learning higher-level features with convolutional restricted Boltzmann machines for sentiment analysis

In recent years, learning word vector representations has attracted much interest in Natural Language Processing. Word representations or embeddings learned using unsupervised methods help addressing the problem of traditional bag-of-word approaches which fail to capture contextual semantics. In this paper we go beyond the vector representations at the word level and propose a novel framework that learns higher-level feature representations of n-grams, phrases and sentences using a deep neural network built from stacked Convolutional Restricted Boltzmann Machines (CRBMs). These representations have been shown to map syntactically and semantically related n-grams to closeby locations in the hidden feature space. We have experimented to additionally incorporate these higher-level features into supervised classifier training for two sentiment analysis tasks: subjectivity classification and sentiment classification. Our results have demonstrated the success of our proposed framework with 4% improvement in accuracy observed for subjectivity classification and improved the results achieved for sentiment classification over models trained without our higher level features.

Learning from Geometry

Subspaces and manifolds are two powerful models for high dimensional signals. Subspaces model linear correlation and are a good fit to signals generated by physical systems, such as frontal images of human faces and multiple sources impinging at an antenna array. Manifolds model sources that are not linearly correlated, but where signals are determined by a small number of parameters. Examples are images of human faces under different poses or expressions, and handwritten digits with varying styles. However, there will always be some degree of model mismatch between the subspace or manifold model and the true statistics of the source. This dissertation exploits subspace and manifold models as prior information in various signal processing and machine learning tasks.

A near-low-rank Gaussian mixture model measures proximity to a union of linear or affine subspaces. This simple model can effectively capture the signal distribution when each class is near a subspace. This dissertation studies how the pairwise geometry between these subspaces affects classification performance. When model mismatch is vanishingly small, the probability of misclassification is determined by the product of the sines of the principal angles between subspaces. When the model mismatch is more significant, the probability of misclassification is determined by the sum of the squares of the sines of the principal angles. Reliability of classification is derived in terms of the distribution of signal energy across principal vectors. Larger principal angles lead to smaller classification error, motivating a linear transform that optimizes principal angles. This linear transformation, termed TRAIT, also preserves some specific features in each class, being complementary to a recently developed Low Rank Transform (LRT). Moreover, when the model mismatch is more significant, TRAIT shows superior performance compared to LRT.

The manifold model enforces a constraint on the freedom of data variation. Learning features that are robust to data variation is very important, especially when the size of the training set is small. A learning machine with large numbers of parameters, e.g., deep neural network, can well describe a very complicated data distribution. However, it is also more likely to be sensitive to small perturbations of the data, and to suffer from suffer from degraded performance when generalizing to unseen (test) data.

From the perspective of complexity of function classes, such a learning machine has a huge capacity (complexity), which tends to overfit. The manifold model provides us with a way of regularizing the learning machine, so as to reduce the generalization error, therefore mitigate overfiting. Two different overfiting-preventing approaches are proposed, one from the perspective of data variation, the other from capacity/complexity control. In the first approach, the learning machine is encouraged to make decisions that vary smoothly for data points in local neighborhoods on the manifold. In the second approach, a graph adjacency matrix is derived for the manifold, and the learned features are encouraged to be aligned with the principal components of this adjacency matrix. Experimental results on benchmark datasets are demonstrated, showing an obvious advantage of the proposed approaches when the training set is small.

Stochastic optimization makes it possible to track a slowly varying subspace underlying streaming data. By approximating local neighborhoods using affine subspaces, a slowly varying manifold can be efficiently tracked as well, even with corrupted and noisy data. The more the local neighborhoods, the better the approximation, but the higher the computational complexity. A multiscale approximation scheme is proposed, where the local approximating subspaces are organized in a tree structure. Splitting and merging of the tree nodes then allows efficient control of the number of neighbourhoods. Deviation (of each datum) from the learned model is estimated, yielding a series of statistics for anomaly detection. This framework extends the classical {\em changepoint detection} technique, which only works for one dimensional signals. Simulations and experiments highlight the robustness and efficacy of the proposed approach in detecting an abrupt change in an otherwise slowly varying low-dimensional manifold.

CHARACTERIZATION AND THERMAL-HYDRAULIC MODELING OF ENGINEERED BARRIERS FOR A CANADIAN DEEP GEOLOGICAL REPOSITORY FOR USED NUCLEAR FUEL

Within Canada there are more than 2.5 million bundles of spent nuclear fuel with another approximately 2 million bundles to be generated in the future. Canada, and every country around the world that has taken a decision on management of spent nuclear fuel, has decided on long-term containment and isolation of the fuel within a deep geological repository. At depth, a deep geological repository consists of a network of placement rooms where the bundles will be located within a multi-layered system that incorporates engineered and natural barriers. The barriers will be placed in a complex thermal-hydraulic-mechanical-chemical-biological (THMCB) environment. A large database of material properties for all components in the repository are required to construct representative models. Within the repository, the sealing materials will experience elevated temperatures due to the thermal gradient produced by radioactive decay heat from the waste inside the container. Furthermore, high porewater pressure due to the depth of repository along with possibility of elevated salinity of groundwater would cause the bentonite-based materials to be under transient hydraulic conditions. Therefore it is crucial to characterize the sealing materials over a wide range of thermal-hydraulic conditions. A comprehensive experimental program has been conducted to measure properties (mainly focused on thermal properties) of all sealing materials involved in Mark II concept at plausible thermal-hydraulic conditions. The thermal response of Canada’s concept for a deep geological repository has been modelled using experimentally measured thermal properties. Plausible scenarios are defined and the effects of these scenarios are examined on the container surface temperature as well as the surrounding geosphere to assess whether they meet design criteria for the cases studied. The thermal response shows that if all the materials even being at dried condition, repository still performs acceptably as long as sealing materials remain in contact.

The influence of surface and deep cues on primary and secondary school students’ assessment of relevance in Web menus

International audience