Food recognition using neural network classifier and multiple hypotheses image segmentation

ABSTRACT

This paper proposes the multiple-hypotheses image segmentation and feed-forward neural network classifier for food recognition to improve the performance. Initially, the food or meal image is given as input. Then, the segmentation is applied to identify the regions, where a particular food item is located using salient region detection, multi-scale segmentation, and fast rejection. Then, the features of every food item are extracted by the global feature and local feature extraction. After the features are obtained, the classification is performed for each segmented region using a feed-forward neural network model. Finally, the calorie value is computed with the aid of (i) food volume and (ii) calorie and nutrition measure based on mass value. The experimental results and performance evaluation are validated. The outcome of the proposed method attains 0.947 for Macro Average Accuracy (MAA) and 0.959 for Standard Accuracy (SA), which provides better classification performance.     

Image Super-Resolution Based on Generative Adversarial Networks: A Brief Review

Single image super resolution (SISR) is an important research content in the field of computer vision and image processing. With the rapid development of deep neural networks, different image super-resolution models have emerged. Compared to some traditional SISR methods, deep learning-based methods can complete the superresolution tasks through a single image. In addition, compared with the SISR methods using traditional convolutional neural networks, SISR based on generative adversarial networks (GAN) has achieved the most advanced visual performance. In this review, we first explore the challenges faced by SISR and introduce some common datasets and evaluation metrics. Then, we review the improved network structures and loss functions of GAN-based perceptual SISR. Subsequently, the advantages and disadvantages of different networks are analyzed by multiple comparative experiments. Finally, we summarize the paper and look forward to the future development trends of GAN-based perceptual SISR.     

PLC-RF diversity: channel outage analysis

Telecommunication Systems - Signals from indoor access points in a wireless fidelity suffer from high attenuation as a result of the thick walls and metal structures in the multilevel-storey...     

Preparation of Pure Uranium,Thorium, and Yttrium Oxides from El-Garra El-Hamra Sulfate Leach Liquor

Radiochemistry - The study deals with separation of uranium, thorium, and yttrium with highest possible purity from El-Garaa El-Hamara sulfate leach liquor. From the latter, almost complete...     

Analysis of phosphorus removal and anaerobic selector performance in a full-scale activated sludge process in Singapore.

This paper analyses the performance of the anaerobic selector (A/O process) in a full-scale activated sludge process receiving mostly industrial sewage discharge (> 60%) in Singapore. In addition to the sludge settleability, enhanced biological phosphorus removal (EBPR) was studied. The sludge volume index (SVI) reduced from 200 to 80 ml g(-1) and foaming was suppressed significantly, indicating the effectiveness of the anaerobic selector in improving sludge settleability. The phosphorus removal efficiency was 66%, and 7.5 mg HAc-COD was consumed per mg PO4(3-) -P removed. In the anaerobic compartment, 31% of the SCOD and 73% of the acetic acid in the settled sewage were removed with PO4(3-) -P release of 14.1 mg PO4(3-)-P l(-1). The linear correlation between PO4(3-) -P release in the anaerobic compartment and PO4(3-) -P uptake in the aerobic compartment indicates that there is about 0.8 mg PO4(3-) -P release in the anaerobic compartment per mg PO34(3-) -P uptake in the aerobic compartment. The fates of volatile fatty acids (VFAs) and its short chain acids (SCAs) in the process were studied and discussed.     

Using decision analysis to determine optimal experimental design for monitoring sewer exfiltration with tracers.

The tracer methods developed to assess exfiltration from sewers in the European project APUSS (assessment of the performance of sewer systems) have a high degree of freedom with regard to the choice of tracer and the dosing strategy. These can lead to very different degrees of uncertainty in the measured exfiltration ratio. In this study, we demonstrate how to select an optimal experimental design using decision analysis, which accounts for this uncertainty and its associated costs. Although the results are site-specific, we can conclude generally that, when NaCl is used as the tracer, the accuracy of the exfiltration estimate is most sensitive to the amount of tracer used and the starting time of the experiment.     

Tracking influent inorganic suspended solids through wastewater treatment plants.

From an experimental and theoretical investigation of the continuity of influent inorganic suspended solids (ISS) along the links connecting the primary settling tank (PST), fully aerobic or N removal activated sludge (AS) and anaerobic and aerobic sludge digestion unit operations, it was found that the influent wastewater (fixed) ISS concentration is conserved through primary sludge anaerobic digestion, activated sludge and aerobic digestion unit operations. However, the measured ISS flux at different stages through a series of wastewater treatment plant (WWTP) unit operations is not equal to the influent ISS flux, because the ordinary heterotrophic organisms (OHO) biomass contributes to the ISS flux by differing amounts depending on the active fraction of the VSS solids at that stage.     

Sampling for representative micropollutant loads in sewer systems.

Most commercially available auto-sampling devices do not support a continuous flow-proportional sampling mode, which would conceptually be the best for collecting representative composite samples. Instead different discrete sampling modes are available. Household chemicals can show considerable random short-term variations. With the anticorrosive benzotriazole, relating to a middle-frequent household activity, we show that, besides an accurate flow meter, mainly three factors are decisive for the representativeness of a substance's average load: the substance's load pattern, the sampling frequency and the length of the composite sample. When the sampling intervals are 10 minutes or longer, errors in the order of +/-40% (standard deviation) or more have to be accepted, if the substance of interest is contained in a low number of wastewater pulses (i.e., the level of household activity). This particularly holds true for specific pharmaceuticals e.g. carbamazepine. Ammonium would be less critical, because it relates to a larger number of sources in the same catchment.     

The Nexus of Corporate Entrepreneurship and Radical Innovation

This paper explores the linkage between the entrepreneurial orientation of established firms and the development of radical innovation. Through five case studies in firms involved in radical innovation, three propositions are developed, suggesting that proactiveness, risk‐taking and autonomy stimulate the development of radical innovation, whereas competitive aggressiveness does not necessarily do so, as radical innovations are directed towards the creation of entirely new arenas of business, where existing competitors are not present.     

Aripiprazole as a potential pharmacotherapy for stimulant dependence: Human laboratory studies with d-amphetamine.

Amphetamine and cocaine dependence present significant public health concerns, yet no broadly effective pharmacotherapy for stimulant dependence has been developed. Two human laboratory studies are reviewed that tested the ability of aripiprazole, a novel antipsychotic with partial agonist activity at D2 dopamine receptors, to alter the behavioral effects of stimulants using d-amphetamine as a model agent. In each of these experiments, volunteers learned to discriminate 15 mg d-amphetamine (i.e., ≥80% drug-appropriate responding over 4 consecutive sessions). The effects of a range of doses of d-amphetamine (0, 2.5, 5, 10, and 15 mg) were then tested alone and following pretreatment with aripiprazole (20 mg in Experiment 1; 10 mg in Experiment 2). In Experiment 1, aripiprazole (20 mg) attenuated the discriminative stimulus and many of the subject-rated effects of amphetamine. Aripiprazole alone produced performance decrements. To determine whether a lower dose of aripiprazole would also attenuate the behavioral effects of d-amphetamine without impairing performance, Experiment 2 was conducted. Aripiprazole (10 mg) failed to alter the discriminative-stimulus effects but attenuated some of the subject-rated effects of d-amphetamine. This dose of aripiprazole did not impair performance. The results of these experiments indicate that aripiprazole may have clinical utility in treating stimulant dependence. Future human laboratory research should better model the clinical use of aripiprazole by examining the effects of chronic aripiprazole combined with either methamphetamine or cocaine in dependent individuals. A large-scale clinical trial is also needed to evaluate the efficacy of aripiprazole for the treatment of stimulant dependence. (PsycINFO Database Record (c) 2010 APA, all rights reserved)