Underwater image super-resolution using multi-stage information distillation networks

Recently, single image super-resolution (SISR) has been widely applied in the fields of underwater robot vision and obtained remarkable performance. However, most current methods generally suffered from the problem of a heavy burden on computational resources with large model sizes, which limited their real-world underwater robotic applications. In this paper, we introduce and tackle the super resolution (SR) problem for underwater robot vision and provide an efficient solution for near real-time applications. We present a novel lightweight multi-stage information distillation network, named MSIDN, for better balancing performance against applicability, which aggregates the local distilled features from different stages for more powerful feature representation. Moreover, a novel recursive residual feature distillation (RRFD) module is constructed to progressively extract useful features with a modest number of parameters in each stage. We also propose a channel interaction & distillation (CI&D) module that employs channel split operation on the preceding features to produce two-part features and utilizes the inter channel-wise interaction information between them to generate the distilled features, which can effectively extract the useful information of current stage without extra parameters. Besides, we present USR-2K dataset, a collection of over 1.6K samples for large-scale underwater image SR training, and a testset with an additional 400 samples for benchmark evaluation. Extensive experiments on several standard benchmark datasets show that the proposed MSIDN can provide state-of-the-art or even better performance in both quantitative and qualitative measurements.     

WavNet — Visual saliency detection using Discrete Wavelet Convolutional Neural Network

In the recent advancements in image and video analysis, the detection of salient regions in the image becomes the initial step. This plays a crucial role in deciding the performance of such algorithms. In this work, a Multi-Resolution Feature Extraction (MRFE) technique that makes use of Discrete Wavelet Convolutional Neural Network (DWCNN) for generating features is employed. An Enhanced Feature Extraction (EFE) module extracts additional features from the high level features of the DWCNN, which are used to frame both channel as well as spatial attention models for yielding contextual attention maps. A new hybrid loss function is also proposed, which is a combination of Balanced Cross Entropy (BCE) loss and Edge based Structural Similarity (ESSIM) loss that effectively identifies and segments the salient regions with clear boundaries. The method is tested exhaustively with five different benchmark datasets and is proved superior to the existing state-of-the-art methods with a minimum Mean Absolute error (MAE) of 0.03 and F-measure of 0.956.     

Modeling Sialidosis with Neural Precursor Cells Derived from Patient-Derived Induced Pluripotent Stem Cells

Sialidosis, caused by a genetic deficiency of the lysosomal sialidase gene (NEU1), is a systemic disease involving various tissues and organs, including the nervous system. Understanding the neurological dysfunction and pathology associated with sialidosis remains a challenge, partially due to the lack of a human model system. In this study, we have generated two types of induced pluripotent stem cells (iPSCs) with sialidosis-specific NEU1^G227R and NEU1^V275A/R347Q mutations (sialidosis-iPSCs), and further differentiated them into neural precursor cells (iNPCs). Characterization of NEU1^G227R- and NEU1^V275A/R347Q- mutated iNPCs derived from sialidosis-iPSCs (sialidosis-iNPCs) validated that sialidosis-iNPCs faithfully recapitulate key disease-specific phenotypes, including reduced NEU1 activity and impaired lysosomal and autophagic function. In particular, these cells showed defective differentiation into oligodendrocytes and astrocytes, while their neuronal differentiation was not notably affected. Importantly, we found that the phenotypic defects of sialidosis-iNPCs, such as impaired differentiation capacity, could be effectively rescued by the induction of autophagy with rapamycin. Our results demonstrate the first use of a sialidosis-iNPC model with NEU1^G227R- and NEU1^V275A/R347Q- mutation(s) to study the neurological defects of sialidosis, particularly those related to a defective autophagy–lysosome pathway, and may help accelerate the development of new drugs and therapeutics to combat sialidosis and other LSDs.     

Exploration of GH94 Sequence Space for Enzyme Discovery Reveals a Novel Glucosylgalactose Phosphorylase Specificity

The substantial increase in DNA sequencing efforts has led to a rapid expansion of available sequences in glycoside hydrolase families. The ever-increasing sequence space presents considerable opportunities for the search for enzymes with novel functionalities. In this work, the sequence-function space of glycoside hydrolase family 94 (GH94) was explored in detail, using a combined approach of phylogenetic analysis and sequence similarity networks. The identification and experimental screening of unknown clusters led to the discovery of an enzyme from the soil bacterium Paenibacillus polymyxa that acts as a 4-O-β-d -glucosyl-d -galactose phosphorylase (GGalP), a specificity that has not been reported to date. Detailed characterization of GGalP revealed that its kinetic parameters were consistent with those of other known phosphorylases. Furthermore, the enzyme could be used for production of the rare disaccharides 4-O-β-d -glucosyl-d -galactose and 4-O-β-d -glucosyl-l -arabinose. Our current work highlights the power of rational sequence space exploration in the search for novel enzyme specificities, as well as the potential of phosphorylases for rare disaccharide synthesis.     

Transmission line detection in aerial images: An instance segmentation approach based on multitask neural networks

Camera-based transmission line detection (TLD) is a fundamental and crucial task for automatically patrolling powerlines by aircraft. Motivated by instance segmentation, a TLD algorithm is proposed in this paper with a novel deep neural network, i.e., CableNet. The network structure is designed based on fully convolutional networks (FCNs) with two major improvements, considering the specific appearance characteristics of transmission lines. First, overlaying dilated convolutional layers and spatial convolutional layers are configured to better represent continuous long and thin cable shapes. Second, two branches of outputs are arranged to generate multidimensional feature maps for instance segmentation. Thus, cable pixels can be detected and assigned cable IDs simultaneously. Multiple experiments are conducted on aerial images, and the results show that the proposed algorithm obtains reliable detection performance and is superior to traditional TLD methods. Meanwhile, segmented pixels can be accurately identified as cable instances, contributing to line fitting for further applications.     

一类分数阶微积分的中值定理及应用

Riemann-Liouville分数阶微积分算子是一类带有一个函数的分数阶微积分算子的特殊情形，以Riemann-Liouville分数阶微积分算子的积分中值定理和微分中值定理为基础，我们得到了一类带有一个函数的分数阶微积分算子的积分中值定理和微分中值定理，并给出其在计算方面的一些应用。     

基于GA-BP的汽车风振噪声声品质预测模型

目前对于汽车风振噪声的优化研究主要以声压级（Sound pressure level，SPL）作为单一评价指标，既不能全面反映噪声的物理属性，也无法考虑人耳对噪声的主观认知过程。为准确评价风振噪声，引入声品质，运用大涡模拟（Large eddy simulation，LES）对风振噪声进行数值仿真，根据实车道路试验判断仿真的准确性；对仿真结果进行声品质客观评价与主观评价，综合声品质客观评价参数与声品质主观评价试验结果建立BP神经网络预测模型；利用遗传算法（Genetic algorithm，GA），进一步对BP神经网络的结构参数进行优化，建立GA-BP声品质预测模型。研究结果表明，GA-BP声品质预测模型在训练速度和预测精度上都优于BP神经网络预测模型。预测模型基于声品质主客观评价结果，其预测值可以代替传统的声压级评价指标，为风振噪声提供更为准确合理的评价。     

基于NR-Transformer的集群作业运行时间预测

高性能集群的作业调度通常使用作业调度系统来实现，准确填写作业运行时间能在很大程度上提升作业调度效率。现有的研究通常使用机器学习的预测方式，在预测精度和实用性上还存在一定的提升空间。为了进一步提高集群作业运行时间预测的准确率，考虑先对集群作业日志进行聚类，将作业类别信息添加到作业特征中，再使用基于注意力机制的NR-Transformer网络对作业日志数据建模和预测。在数据处理上，根据与预测目标的相关性、特征的完整性和数据的有效性，从历史日志数据集中筛选出7维特征，并按作业运行时间的长度将其划分为多个作业集，再对各作业集分别进行训练和预测。实验结果表明，相比于传统机器学习和BP神经网络，时序神经网络结构有更好的预测性能，其中NR-Transformer在各作业集上都有较好的性能。     

基于音频特征的工程车辆工况识别研究

高效率地使用工程车辆是工程项目管理中节约成本的有效方法，无人监管环境下工程车辆的工况识别，是实现工程车辆高效率使用的有效手段。目前以GPS等技术为核心的车辆智能管理系统未对工程车辆进行工况识别，提出一种基于GRU循环神经网络的工程车辆工况识别方法，通过对工程车辆在不同工况下产生的音频信号进行分析，从中提取Mel倒谱系数作为主要特征，构建GRU循环神经网络模型进行训练和识别。实验结果表明，该方法可以实现对工程车辆工况的有效识别。