The use of microtechnology’s in the construction of water aeration installations

Water aeration systems are highly efficient if the dispersion of air in the water is carried out in a controlled and uniform manner. The use of fine bubble generators ensures this and in addition, creates a small loss of pressure when air passes through them. The paper demonstrates that producing as few air bubbles as possible leads to a more efficient aeration process. Two water aeration installations are compared: - The first has a perforated plate with 152 orifices Ø 0.1 mm; - The second has four perforated plates, each with 113 orifices Ø 0.05 mm; Both installations are successively supplied with the same flow rate of compressed air, at the same temperature and at the same initial dissolved oxygen concentration in the water.

Early Labeled and Small Loss Selection Semi-Supervised Learning Method for Remote Sensing Image Scene Classification

The classification of aerial scenes has been extensively studied as the basic work of remote sensing image processing and interpretation. However, the performance of remote sensing image scene classification based on deep neural networks is limited by the number of labeled samples. In order to alleviate the demand for massive labeled samples, various methods have been proposed to apply semi-supervised learning to train the classifier using labeled and unlabeled samples. However, considering the complex contextual relationship and huge spatial differences, the existing semi-supervised learning methods bring different degrees of incorrectly labeled samples when pseudo-labeling unlabeled data. In particular, when the number of labeled samples is small, it affects the generalization performance of the model. In this article, we propose a novel semi-supervised learning method with early labeled and small loss selection. First, the model learns the characteristics of simple samples in the early stage and uses multiple early models to screen out a small number of unlabeled samples for pseudo-labeling based on this characteristic. Then, the model is trained in a semi-supervised manner by combining labeled samples, pseudo-labeled samples, and unlabeled samples. In the training process of the model, small loss selection is used to further eliminate some of the noisy labeled samples to improve the recognition accuracy of the model. Finally, in order to verify the effectiveness of the proposed method, it is compared with several state-of-the-art semi-supervised classification methods. The results show that when there are only a few labeled samples in remote sensing image scene classification, our method is always better than previous methods.

Towards Understanding Deep Learning from Noisy Labels with Small-Loss Criterion

Deep neural networks need large amounts of labeled data to achieve good performance. In real-world applications, labels are usually collected from non-experts such as crowdsourcing to save cost and thus are noisy. In the past few years, deep learning methods for dealing with noisy labels have been developed, many of which are based on the small-loss criterion. However, there are few theoretical analyses to explain why these methods could learn well from noisy labels. In this paper, we theoretically explain why the widely-used small-loss criterion works. Based on the explanation, we reformalize the vanilla small-loss criterion to better tackle noisy labels. The experimental results verify our theoretical explanation and also demonstrate the effectiveness of the reformalization.

What Neutrino Reaction Has a Large Cross-Section?

All reactions that are currently used to detect neutrinos are endothermic (more precisely, endo-energy). They occur at the expense of the energy of neutrino that initiates them. These reactions are characterized by a very small cross-section, which is close in magnitude to the 10−20 barn. Beta decay is an exo-thermal (more precisely, exo-energetic) reaction. Currently, it seems that the entire physical community believes that the beta decay phenomenon occurs completely by accident. However, recent experiments with reactor neutrinos [1, 2] have shown that their flux makes an additional contribution to the beta decay rate. Since beta decay is an exo-energetic reaction, neutrinos catalyze beta-active nuclei without losing own energy (or with a small loss of it). The cross-section of this process is much larger than the cross-section of the endo-energetic interaction of neutrinos with matter. Experimental measurements show that the cross section of reactor neutrinos with 63Ni nuclei is close to 1 barn.

Traumatic Brain Injury in Children

Traumatic brain injury (TBI) in children occurs as a result of a sudden bump, roll, or jerk to the head or a penetrating injury to the head that interferes the normal brain function. Traumatic brain injury (TBI) is the leading cause of death and disability in children. More than half a million children present annually to the emergency department for TBI-related visits, and resulting in the death of >7,000 children annually in the United States, with highest incident rates seen in children aged 0–4 years and adolescents aged 15 to 19 years. In Indonesia, from Riskesdas data in 2013 shows the incidence of head trauma in children is about 0.5% of the population from other injury rates. Pediatric TBI is associated with an array of negative outcomes, including impaired cognitive and academic abilities, social impairments, and behavioral problems. The scalp is highly vascularized and a potential cause of lethal blood loss. Even a small loss of blood volume can lead to hemorrhagic shock in a newborn, infant, and toddler, which may occur without apparent external bleeding.

Diffeomorphism cocycles over partially hyperbolic systems

We consider Hölder continuous cocycles over an accessible partially hyperbolic system with values in the group of diffeomorphisms of a compact manifold $\mathcal {M}$ . We obtain several results for this setting. If a cocycle is bounded in $C^{1+\gamma }$ , we show that it has a continuous invariant family of $\gamma $ -Hölder Riemannian metrics on $\mathcal {M}$ . We establish continuity of a measurable conjugacy between two cocycles assuming bunching or existence of holonomies for both and pre-compactness in $C^0$ for one of them. We give conditions for existence of a continuous conjugacy between two cocycles in terms of their cycle weights. We also study the relation between the conjugacy and holonomies of the cocycles. Our results give arbitrarily small loss of regularity of the conjugacy along the fiber compared to that of the holonomies and of the cocycle.

Immunogenicity of reduced-dose monovalent type 2 oral poliovirus vaccine in Mocuba, Mozambique

Background Monovalent type 2 oral poliovirus vaccine (mOPV2) stockpile is low. One potential strategy to stretch the existing mOPV2 supply is to administer a reduced dose: one-drop instead of two-drops. Methods We conducted a randomized, controlled, open-label, non-inferiority trial (10% margin) to compared immunogenicity following administration of one versus two-drops of mOPV2. We enrolled 9-22-months old infants from Mocuba district of Mozambique. Poliovirus neutralizing antibodies were measured in sera collected before and one month after mOPV2 administration. Immune response was defined as seroconversion from seronegative (<1:8) at baseline to seropositive (>1:8) after vaccination or boosting titers by >4-fold for those with titers between 1:8 and 1:362 at baseline. The trial was registered at anzctr.org.au (number ACTRN12619000184178p). Results We enrolled 378 children and 262 (69%) completed per-protocol requirements. Immune response of mOPV2 was 53.6% (95% confidence interval [CI]: 44.9%-62.1%) and 60.6% (95% CI: 52.2%-68.4%) in 1-drop and 2-drops recipients, respectively. The non-inferiority margin of the 10% was not reached (difference=7.0%; 95%CI= -5.0-19.0). Conclusion A small loss of immunogenicity of reduced mOPV2 was observed. Although the non-inferiority target was not achieved, the Strategic Advisory Group of Experts on Immunization, recommended the 1-drop strategy as a dose-sparing measure if mOPV2 supplies deteriorate further.

The Horn Technique for Nasal Tip Support in Rhinoplasty

Introduction The predictability of nasal tip projection and rotation after aesthetic surgery is a challenge. Tongue-in-groove (TIG) is an effective technique to control tip projection and rotation, but there may be a small loss of projection and rotation of the tip lobe due to lack of support between the anterior septal angle and the domus, since this region is sustained by medial crusts suture-linked and interdomus sutures. Objective To describe a new surgery technique in an attempt to correct the lack of support for the nasal tip after lowering the nasal dorsum. Methods The horn technique consists in preserving a square of cartilage during the removal of the nasal dorsum and septum excess in patients with long and projected nose. This piece will give greater support to the TIG technique and greater predictability of the rotation and projection of the nasal tip. Results Between 2016 and 2018, 50 patients with long and projected noses were submitted to the “horn technique” surgery. They were submitted to the TIG technique associated to the horn technique. A retrospective review of the preoperative and postoperative photographs (3 months to 1 year) of these patients treated with the horn technique were analyzed and showed better support of the nasal tip. Conclusion The horn technique provides greater support to the projection and rotation of rhinoplasties in patients with long and projected nose.