Six Undescribed Capnosane-Type Macrocyclic Diterpenoids from South China Sea Soft Coral Sarcophyton crassocaule: Structural Determination and Biological Evaluation.

Six undescribed capnosane-type macrocyclic diterpenes sarcocrassolins A-F (1-6) and one related known analog pavidolide D (7) were isolated from Sarcophyton crassocaule, a soft coral collected off the Nansha Islands, in the South China Sea. Their complete structures, relative configurations and absolute configurations were established through comprehensive spectroscopic analysis, quantum mechanical nuclear magnetic resonance (QM-NMR) and single-crystal X-ray diffraction. Sarcocrassolins D (4) and E (5) showed inhibitory activity against lipopolysaccharide (LPS)-stimulated inflammatory responses in RAW264.7 cells with IC50 values of 76.8 ± 8.0 µM and 93.0 ± 3.8 µM, respectively.

p-Terphenyl and Diphenyl Ether Derivatives from the Marine-Derived Fungus Aspergillus candidus HM5-4.

Two undescribed p-terphenyl derivatives, asperterphenylcins A-B (1-2), and two undescribed diphenyl ether derivatives, asperdiphenylcins A-B (3-4), together with three previously described p-terphenyl derivatives-4″-deoxyterprenin (5), terphenyllin (6), and 3″-hydroxyterphenyllin (7)-were obtained from the solid-rice culture of the marine-derived fungus Aspergillus candidus HM5-4, which was isolated from sponges from the South China Sea. Their structures were elucidated by HRESIMS data and NMR spectroscopic analysis. Compound 1 showed a strong inhibitory effect on Neoscytalidium dimidiatum, with an inhibition circle diameter of 31.67 ± 2.36 mm at a concentration of 10.0 µg/disc. Compounds 5 and 7 displayed cytotoxic activity against human chronic myeloid leukemia cells (K562), human liver cancer cells (BEL-7402), human gastric cancer cells (SGC-7901), human non-small cell lung cancer cells (A549) and human HeLa cervical cancer cells, with IC50 values ranging from 3.32 to 60.36 µM, respectively. Compounds 2, 6 and 7 showed potent inhibitory activity against α-glucosidase, with IC50 values of 1.26 ± 0.19, 2.16 ± 0.44 and 13.22 ± 0.55 µM, respectively.

A Systematic IoU-Related Method: Beyond Simplified Regression for Better Localization.

Four-variable-independent-regression localization losses, such as Smooth- l1 Loss, are used by default in modern detectors. Nevertheless, this kind of loss is oversimplified so that it is inconsistent with the final evaluation metric, intersection over union (IoU). Directly employing the standard IoU is also not infeasible, since the constant-zero plateau in the case of non-overlapping boxes and the non-zero gradient at the minimum may make it not trainable. Accordingly, we propose a systematic method to address these problems. Firstly, we propose a new metric, the extended IoU (EIoU), which is well-defined when two boxes are not overlapping and reduced to the standard IoU when overlapping. Secondly, we present the convexification technique (CT) to construct a loss on the basis of EIoU, which can guarantee the gradient at the minimum to be zero. Thirdly, we propose a steady optimization technique (SOT) to make the fractional EIoU loss approaching the minimum more steadily and smoothly. Fourthly, to fully exploit the capability of the EIoU based loss, we introduce an interrelated IoU-predicting head to further boost localization accuracy. With the proposed contributions, the new method incorporated into Faster R-CNN with ResNet50+FPN as the backbone yields 4.2 mAP gain on VOC2007 and 2.3 mAP gain on COCO2017 over the baseline Smooth- l1 Loss, at almost no training and inferencing computational cost. Specifically, the stricter the metric is, the more notable the gain is, improving 8.2 mAP on VOC2007 and 5.4 mAP on COCO2017 at metric AP90 .

Discriminative Feature Selection via Employing Smooth and Robust Hinge Loss.

A wide variety of sparsity-inducing feature selection methods have been developed in recent years. Most of the loss functions of these approaches are built upon regression since it is general and easy to optimize, but regression is not well suitable for classification. In contrast, the hinge loss (HL) of support vector machines has proved to be powerful to handle classification tasks, but a model with existing multiclass HL and sparsity regularization is difficult to optimize. In view of that, we propose a new loss, called smooth and robust HL, which gathers the merits of regression and HL but overcome their drawbacks, and apply it to our sparsity regularized feature selection model. To optimize the model, we present a new variant of accelerated proximal gradient (APG) algorithm, which boosts the discriminative margins among different classes, compared with standard APG algorithms. We further propose an efficient optimization technique to solve the proximal projection problem at each iteration step, which is a key component of the new APG algorithm. We theoretically prove that the new APG algorithm converges at rate O(1/k2) if it is convex ( k is the iteration counter), which is the optimal convergence rate for smooth problems. Experimental results on nine publicly available data sets demonstrate the effectiveness of our method.

Feature selection by optimizing a lower bound of conditional mutual information.

A unified framework is proposed to select features by optimizing computationally feasible approximations of high-dimensional conditional mutual information (CMI) between features and their associated class label under different assumptions. Under this unified framework, state-of-the-art information theory based feature selection algorithms are rederived, and a new algorithm is proposed to select features by optimizing a lower bound of the CMI with a weaker assumption than those adopted by existing methods. The new feature selection method integrates a plug-in component to distinguish redundant features from irrelevant ones for improving the feature selection robustness. Furthermore, a novel metric is proposed to evaluate feature selection methods based on simulated data. The proposed method has been compared with state-of-the-art feature selection methods based on the new evaluation metric and classification performance of classifiers built upon the selected features. The experiment results have demonstrated that the proposed method could achieve promising performance in a variety of feature selection problems.

Improvement of energy efficiency via spectrum optimization of excitation sequence for multichannel simultaneously triggered airborne sonar system.

Both the energy efficiency and correlation characteristics are important in airborne sonar systems to realize multichannel ultrasonic transducers working together. High energy efficiency can increase echo energy and measurement range, and sharp autocorrelation and flat cross correlation can help eliminate cross-talk among multichannel transducers. This paper addresses energy efficiency optimization under the premise that cross-talk between different sonar transducers can be avoided. The nondominated sorting genetic algorithm-II is applied to optimize both the spectrum and correlation characteristics of the excitation sequence. The central idea of the spectrum optimization is to distribute most of the energy of the excitation sequence within the frequency band of the sonar transducer; thus, less energy is filtered out by the transducers. Real experiments show that a sonar system consisting of eight-channel Polaroid 600 series electrostatic transducers excited with 2 ms optimized pulse-position-modulation sequences can work together without cross-talk and can measure distances up to 650 cm with maximal 1% relative error.