A Survey on Image Semantic Segmentation Using Deep Learning Techniques

Image semantic segmentation is an important branch of computer vision of a wide variety of practical applications such as medical image analysis,autonomous driving,virtual or augmented reality,etc.In recent years,due to the remarkable performance of transformer and multilayer perceptron(MLP)in computer vision,which is equivalent to convolutional neural network(CNN),there has been a substantial amount of image semantic segmentation works aimed at developing different types of deep learning architecture.This survey aims to provide a comprehensive overview of deep learning methods in the field of general image semantic segmentation.Firstly,the commonly used image segmentation datasets are listed.Next,extensive pioneering works are deeply studied from multiple perspectives(e.g.,network structures,feature fusion methods,attention mechanisms),and are divided into four categories according to different network architectures:CNN-based architectures,transformer-based architectures,MLP-based architectures,and others.Furthermore,this paper presents some common evaluation metrics and compares the respective advantages and limitations of popular techniques both in terms of architectural design and their experimental value on the most widely used datasets.Finally,possible future research directions and challenges are discussed for the reference of other researchers.

Gate-Attention and Dual-End Enhancement Mechanism for Multi-Label Text Classification

In the realm of Multi-Label Text Classification(MLTC),the dual challenges of extracting rich semantic features from text and discerning inter-label relationships have spurred innovative approaches.Many studies in semantic feature extraction have turned to external knowledge to augment the model’s grasp of textual content,often overlooking intrinsic textual cues such as label statistical features.In contrast,these endogenous insights naturally align with the classification task.In our paper,to complement this focus on intrinsic knowledge,we introduce a novel Gate-Attention mechanism.This mechanism adeptly integrates statistical features from the text itself into the semantic fabric,enhancing the model’s capacity to understand and represent the data.Additionally,to address the intricate task of mining label correlations,we propose a Dual-end enhancement mechanism.This mechanism effectively mitigates the challenges of information loss and erroneous transmission inherent in traditional long short term memory propagation.We conducted an extensive battery of experiments on the AAPD and RCV1-2 datasets.These experiments serve the dual purpose of confirming the efficacy of both the Gate-Attention mechanism and the Dual-end enhancement mechanism.Our final model unequivocally outperforms the baseline model,attesting to its robustness.These findings emphatically underscore the imperativeness of taking into account not just external knowledge but also the inherent intricacies of textual data when crafting potent MLTC models.

Comparative study of two lattice Boltzmann multiphase models for simulating wetting phenomena: implementing static contact angles based on the geometric formulation

Wetting phenomena are widespread in nature and industrial applications. In general, systems concerning wetting phenomena are typical multicomponent/multiphase complex fluid systems. Simulating the behavior of such systems is important to both scientific research and practical applications. It is challenging due to the complexity of the phenomena and difficulties in choosing an appropriate numerical method. To provide some detailed guidelines for selecting a suitable multiphase lattice Boltzmann model, two kinds of lattice Boltzmann multiphase models, the modified S-C model and the H-C-Z model, are used in this paper to investigate the static contact angle on solid surfaces with different wettability combined with the geometric formulation(Ding, H. and Spelt, P.D. M. Wetting condition in diffuse interface simulations of contact line motion. Physical Review E, 75(4), 046708(2007)). The specific characteristics and computational performance of these two lattice Boltzmann method(LBM) multiphase models are analyzed including relationship between surface tension and the control parameters, the achievable range of the static contact angle, the maximum magnitude of the spurious currents(MMSC), and most importantly, the convergence rate of the two models on simulating the static contact angle. The results show that a wide range of static contact angles from wetting to non-wetting can be realized for both models. MMSC mainly depends on the surface tension. With the numerical parameters used in this work, the maximum magnitudes of the spurious currents of the two models are on the same order of magnitude. MMSC of the S-C model is universally larger than that of the H-C-Z model. The convergence rate of the S-C model is much faster than that of the H-C-Z model. The major foci in this work are the frequently-omitted important details in simulating wetting phenomena. Thus, the major findings in this work can provide suggestions for simulating wetting phenomena with LBM multiphase models along with the geometric formulation.

经脐单孔腹腔镜后鞘后入路在全腹膜外腹股沟疝修补术中的应用

目的探讨经脐单孔腹腔镜后鞘后入路在全腹膜外疝修补术(TEP)中的临床应用。方法选取2022年01月至2023年7月80例腹股沟疝患者,按照数字表法将患者随机分为实验组(行经脐后鞘后入路单孔TEP)和对照组(行常规TEP),每组各40例。采用 SPSS 22.0软件处理数据,手术相关指标等计量资料以(x±s)表示,采用独立样本t检验;术后并发症、切口满意度等计数资料采用χ^(2)检验。P<0.05 表示为差异有统计学意义。结果实验组患者的术中失血量较少、切口长度较短,住院时间明显缩短,且患者疼痛视觉模拟评分(VAS)评分较低,切口满意度较高,术后并发症较少(P<0.05)。结论经脐单孔腹腔镜后鞘后入路的全腹膜外疝修补术治疗腹股沟疝安全、有效,具有住院时间短、损伤小、并发症少等优点,值得临床上广泛应用。