Semantic Method Research Articles

Ways to translate medical terms into Chinese

The article analyzes the five most commonly used translation methods: transcription, mixed method (transcription and semantic method), semantic method, semantic method with addition of meanings and semantic method with abbreviations. The subject of the article is the translation of the names of foreign medicines in advertising into Chinese. The purpose of the article is to analyze strategies for translating the names of medicines into Chinese in advertising, as well as to identify the most frequent and popular translation methods. The article analyzes the names of foreign medicines (70 names of medicines) and their translation into Chinese. By comparing the frequency of use of each method, the most frequent method is identified when translating foreign drug names into Chinese. In the course of the research, the following methods were used: descriptive method, method of analysis of theoretical literature, as well as methods of systematization and classification of material. The novelty of the study lies in the fact that, to date, the translation of the names of medical terms has not yet been considered within the framework of the theory of translation equivalence. The practical significance of this article lies in the fact that the results of the study can contribute to the further strengthening and development of trade and international communication. The article provides examples illustrating ways to translate drug names into Chinese. During the study, it was revealed that when translating medical terms from foreign languages into Chinese, the most frequent methods are transcription and the semantic method. At the same time, the percentage of use of semantic and mixed methods (which include transcription and semantic method) is also high.The percentage of prevalence of using the semantic method with addition of meanings and abbreviations is significantly lower.

Image Semantic Segmentation Approach for Studying Human Behavior on Image Data

Image semantic segmentation is an essential technique for studying human behavior through image data. This paper proposes an image semantic segmentation method for human behavior research. Firstly, an end-to-end convolutional neural network architecture is proposed, which consists of a depth-separable jump-connected fully convolutional network and a conditional random field network; then jump-connected convolution is used to classify each pixel in the image, and an image semantic segmentation method based on convolutional neural network is proposed; and then a conditional random field network is used to improve the effect of image segmentation of human behavior and a linear modeling and nonlinear modeling method based on the semantic segmentation of conditional random field image is proposed. Finally, using the proposed image segmentation network, the input entrepreneurial image data is semantically segmented to obtain the contour features of the person; and the segmentation of the images in the medical field. The experimental results show that the image semantic segmentation method is effective. It is a new way to use image data to study human behavior and can be extended to other research areas.

Intelligent Identification and Quantitative Characterization of Pores in Shale SEM Images Based on Pore-Net Deep-Learning Network Model

Among the various shale reservoir evaluation methods, the scanning electron microscope (SEM) image method is widely used. Its image can intuitively reflect the development stage of a shale reservoir and is often used for the qualitative characterization of shale pores. However, manual image processing is inefficient and cannot quantitatively characterize pores. The semantic segmentation method of deep learning greatly improves the efficiency of image analysis and can calculate the face rate of shale SEM images to achieve quantitative characterization. In this paper, the high-maturity shale of the Longmaxi Formation in the Changning area of Yibin City, Sichuan Province, and the low-maturity shale of Beibu Gulf Basin in China are studied. Based on the Pore-net network model, the intelligent identification and quantitative characterization of pores in shale SEM images are realized. The pore-net model is improved from the U-net deep-learning network model, which improves the ability of the network model to identify pores. The results show that the pore-net model performs better than the U-net model, FCN model, DeepLab V3 + model, and traditional binarization method. The problem of low accuracy of the traditional pore identification method is solved. The porosity of SEM images of high-maturity shale calculated by the pore-net network model is between 12 and 19% deviation from the experimental data. The calculated porosity of the SEM image of the low-maturity shale has a large deviation from the experimental data, which is between 14 and 47%. Compared with the porosity results calculated by other methods, the results calculated by pore-net are closer to the true value, which proves that the porosity calculated by the pore-net network model is reliable. The deep-learning semantic image segmentation method is suitable for pore recognition of shale SEM images. The fully convolutional neural network model is used to train the manually labeled shale SEM images, which can realize the intelligent recognition and quantitative characterization of the pores in the shale SEM images. It provides a certain reference value for the evaluation of shale oil and gas reservoirs and the study of other porous materials.

대학의 고등교육법 위반사항에 대한 의미연결망 분석

The purpose of this study is to analyze violations of the Higher Education Act by universities, as identified through the internal audits conducted by the Ministry of Education, using semantic network analysis methods. The data analyzed were disclosures related to violations of the Higher Education Act from 2017 to 2022, based on the audit disposition date. Universities with specialized purposes were excluded from the analysis. Keyword frequency analysis and cluster analysis were conducted. Based on the findings of the study, the main results are summarized as follows. Firstly, the keyword frequency analysis encompassing both national and private universities revealed that research, education, corporate, student, hiring, and allowances were the most frequently occurring keywords across the board. Secondly, when analyzed by type of establishment, a key difference between national and private universities was identified: national universities had a higher frequency of student-related keywords compared to private universities, where keywords related to accounting and financial violations were more frequent. Thirdly, the cluster analysis results showed that the violation types were divided into four groups in the overall analysis, categorized as 'Organizational Personnel and Compensation System & Academic System Management,' 'Contract and Accounting Management,' 'Research Management,' and 'Committee and Scholarship System.' Fourthly, when analyzed by type of establishment, national universities were distinctly divided into four groups: research, personnel, academic, and accounting, whereas private universities were categorized into three groups due to research and contract and accounting management being combined into the same group. Based on these findings, recommendations for future research were proposed.

近百年中国法律术语研究

With the accelerated pace of legal globalization, the study of Chinese legal terms has received great attention and promotion. As an important part of legal research and practice, the accuracy and standardization of Chinese legal terms directly affect the effective implementation of laws and the smoothness of international cooperation. This study analyzes the development process, usage context and evolution rules of Chinese legal terms in the last hundred years (from 1923 to 2023), excavates the connotation and evolution path of legal terms, explores the core factors affecting the change of legal terms, and discusses the correlation between the process of Chinese legal reform and the evolution of legal terms. The development of Chinese legal terms is closely related to the reform of China's legal system, which reflects the gradual improvement and internationalization trend of China's legal system. Based on the comprehensive use of comparative legal language and semantics methods, this study predicts the possible trends and challenges of the future development of legal terms, and provides a theoretical basis for the standardization and international exchange of legal terms in China, provides valuable theoretical data for the field of legal linguistics, and expands a new perspective for the study of legal terms.

A semantic big data analysis method based on enhanced neural networks in IoT

AbstractDue to the growth of neural networks, the semantic big data analysis method can classify images at the pixel level, which is very suitable for the needs of IoT. In semantic big data analysis methods, the DeepLab algorithm is an improved and highly accurate algorithm based on enhanced neural networks. However, the DeepLab algorithm does not fully utilize global information, resulting in poor performance for complex scenes. Therefore, this article makes improvements by introducing a global context information module and providing prior information of complex scenes in images. It extracts global information and merges with original features. It improves the expression ability of features. This global context can enhance the accuracy of semantic big data analysis method, and an attention mechanism is designed. The experimental results display that the improved DeepLab semantic big data analysis method based on self‐attention and global context module has good average pixel accuracy and average intersection to union ratio performance on the Pascal VOC 2012 dataset. And the improvement effect is significant.

Coloronyms in the phraseological picture of the world in languages of different structure: linguocultural and pragmatic aspect

Relevance. The relevance of this work is determined by the need for a comprehensive study of the national phraseological picture of the world through cultural, social, pragmatic, geographical factors influencing the transmission of extralingual reality through language nominations, including the expression of an ambivalent idea of the world through understanding the specifics of the lexico-semantic and communicative nature of idioms. Purpose. The purpose of this study is to analyze the parameters and characteristics of phraseological units in the Kazakh and English cultural traditions, to study the lexical-semantic field of coloronyms in the declared languages from the point of view of the communicative nature of these units. Methodology. The method of work is based on a comparative study of phraseological units using linguocultural, communicative-pragmatic, and semantic methods. Results. In this article, a comparative analysis of Kazakh and English phraseological units based on coloronymic vocabulary was made. Linguocultural and pragmatic aspects of the study of idiomatic expressions were considered. An idea was formed about the phraseological picture of the world of Kazakhs and Englishmen. Stylistic parameters, including positive and negative connotations, were evaluated. The ambivalence of stable phrases with a core coloronym was examined. Examples of frequently occurring phrases in Kazakh and English were given. Comparisons were made between different idioms. Additionally, the results of the work were compared with the findings of modern linguists. Conclusions. The study reveals profound cultural, social, and pragmatic influences shaping the phraseological landscape in Kazakh and English, with a specific focus on coloronyms. Through comparative analysis, it elucidates the nuanced ambivalence and stylistic connotations inherent in these idioms. This research enhances our understanding of how color-based phraseological units reflect and convey intricate aspects of lexical semantics and communication in both linguistic traditions. Keywords: idiom; color; semantics; connotation; communicative nature

A deep inverse convolutional neural network-based semantic classification method for land cover remote sensing images

The imbalance of land cover categories is a common problem. Some categories appear less frequently in the image, while others may occupy the vast majority of the proportion. This imbalance can lead the classifier to tend to predict categories with higher frequency of occurrence, while the recognition effect on minority categories is poor. In view of the difficulty of land cover remote sensing image multi-target semantic classification, a semantic classification method of land cover remote sensing image based on depth deconvolution neural network is proposed. In this method, the land cover remote sensing image semantic segmentation algorithm based on depth deconvolution neural network is used to segment the land cover remote sensing image with multi-target semantic segmentation; Four semantic features of color, texture, shape and size in land cover remote sensing image are extracted by using the semantic feature extraction method of remote sensing image based on improved sequential clustering algorithm; The classification and recognition method of remote sensing image semantic features based on random forest algorithm is adopted to classify and identify four semantic feature types of land cover remote sensing image, and realize the semantic classification of land cover remote sensing image. The experimental results show that after this method classifies the multi-target semantic types of land cover remote sensing images, the average values of Dice similarity coefficient and Hausdorff distance are 0.9877 and 0.9911 respectively, which can accurately classify the multi-target semantic types of land cover remote sensing images.

Brain-inspired semantic data augmentation for multi-style images.

Data augmentation is an effective technique for automatically expanding training data in deep learning. Brain-inspired methods are approaches that draw inspiration from the functionality and structure of the human brain and apply these mechanisms and principles to artificial intelligence and computer science. When there is a large style difference between training data and testing data, common data augmentation methods cannot effectively enhance the generalization performance of the deep model. To solve this problem, we improve modeling Domain Shifts with Uncertainty (DSU) and propose a new brain-inspired computer vision image data augmentation method which consists of two key components, namely, using Robust statistics and controlling the Coefficient of variance for DSU (RCDSU) and Feature Data Augmentation (FeatureDA). RCDSU calculates feature statistics (mean and standard deviation) with robust statistics to weaken the influence of outliers, making the statistics close to the real values and improving the robustness of deep learning models. By controlling the coefficient of variance, RCDSU makes the feature statistics shift with semantic preservation and increases shift range. FeatureDA controls the coefficient of variance similarly to generate the augmented features with semantics unchanged and increase the coverage of augmented features. RCDSU and FeatureDA are proposed to perform style transfer and content transfer in the feature space, and improve the generalization ability of the model at the style and content level respectively. On Photo, Art Painting, Cartoon, and Sketch (PACS) multi-style classification task, RCDSU plus FeatureDA achieves competitive accuracy. After adding Gaussian noise to PACS dataset, RCDSU plus FeatureDA shows strong robustness against outliers. FeatureDA achieves excellent results on CIFAR-100 image classification task. RCDSU plus FeatureDA can be applied as a novel brain-inspired semantic data augmentation method with implicit robot automation which is suitable for datasets with large style differences between training and testing data.

A semantic SLAM-based method for navigation and landing of UAVs in indoor environments

Autonomous navigation and landing of Unmanned aerial vehicles (UAVs) are critical functions towards full intelligence in unknown environments. However, current methods heavily rely on positioning devices or geometric maps, leading to two limitations: (1) UAVs fail to work normally indoors when signals are blocked, and (2) Navigation systems are not capable of achieving high-level decision-making missions. In this paper, we propose a novel and complete framework to realize the autonomous landing of UAVs in unknown indoor scenes based on visual SLAM, semantic segmentation, terrain estimation, and a decision-making model. Firstly, our 3D map is built on top of the visual SLAM system with semantic features, and then multiple types of maps (e.g., texture map, octo-map, grid-map, and semantic topology) are created for different function requirements. To achieve high-level scene understanding, we design a data association rule to fuse semantic features extracted by a deep learning model into the mapping process of topology structure. Next, a terrain estimation strategy is performed in the lightweight grid-map, which is modeled only via low-level elevation representation. Through multiple terrain constraints factor, a few optimized landing sites are selected in safe regions via clustering algorithm. Finally, based on the guidance of semantic topology, we construct a decision-making model for UAV landing that effectively integrates terrain safety, environment perception, and path planning. We perform extensive autonomous landing experiments on multiple indoor scenes from the TUM RGB-D dataset, and the full unknown environment maps are created simultaneously. The experiment proves that the proposed framework can select an optimal landing site for the advanced mission of UAVs effectively.

Less Is More: Label Recommendation for Weakly Supervised Point Cloud Semantic Segmentation

Weak supervision has proven to be an effective strategy for reducing the burden of annotating semantic segmentation tasks in 3D space. However, unconstrained or heuristic weakly supervised annotation forms may lead to suboptimal label efficiency. To address this issue, we propose a novel label recommendation framework for weakly supervised point cloud semantic segmentation. Distinct from pre-training and active learning, the label recommendation framework consists of three stages: inductive bias learning, recommendations for points to be labeled, and point cloud semantic segmentation learning. In practice, we first introduce the point cloud upsampling task to induct inductive bias from structural information. During the recommendation stage, we present a cross-scene clustering strategy to generate centers of clustering as recommended points. Then we introduce a recommended point positions attention module LabelAttention to model the long-range dependency under sparse annotations. Additionally, we employ position encoding to enhance the spatial awareness of semantic features. Throughout the framework, the useful information obtained from inductive bias learning is propagated to subsequent semantic segmentation networks in the form of label positions. Experimental results demonstrate that our framework outperforms weakly supervised point cloud semantic segmentation methods and other methods for labeling efficiency on S3DIS and ScanNetV2, even at an extremely low label rate.

SCTNet: Single-Branch CNN with Transformer Semantic Information for Real-Time Segmentation

Recent real-time semantic segmentation methods usually adopt an additional semantic branch to pursue rich long-range context. However, the additional branch incurs undesirable computational overhead and slows inference speed. To eliminate this dilemma, we propose SCTNet, a single branch CNN with transformer semantic information for real-time segmentation. SCTNet enjoys the rich semantic representations of an inference-free semantic branch while retaining the high efficiency of lightweight single branch CNN. SCTNet utilizes a transformer as the training-only semantic branch considering its superb ability to extract long-range context. With the help of the proposed transformer-like CNN block CFBlock and the semantic information alignment module, SCTNet could capture the rich semantic information from the transformer branch in training. During the inference, only the single branch CNN needs to be deployed. We conduct extensive experiments on Cityscapes, ADE20K, and COCO-Stuff-10K, and the results show that our method achieves the new state-of-the-art performance. The code and model is available at https://github.com/xzz777/SCTNet.

Pay Attention to Target: Relation-Aware Temporal Consistency for Domain Adaptive Video Semantic Segmentation

Video semantic segmentation has achieved conspicuous achievements attributed to the development of deep learning, but suffers from labor-intensive annotated training data gathering. To alleviate the data-hunger issue, domain adaptation approaches are developed in the hope of adapting the model trained on the labeled synthetic videos to the real videos in the absence of annotations. By analyzing the dominant paradigm consistency regularization in the domain adaptation task, we find that the bottlenecks exist in previous methods from the perspective of pseudo-labels. To take full advantage of the information contained in the pseudo-labels and empower more effective supervision signals, we propose a coherent PAT network including a target domain focalizer and relation-aware temporal consistency. The proposed PAT network enjoys several merits. First, the target domain focalizer is responsible for paying attention to the target domain, and increasing the accessibility of pseudo-labels in consistency training. Second, the relation-aware temporal consistency aims at modeling the inter-class consistent relationship across frames to equip the model with effective supervision signals. Extensive experimental results on two challenging benchmarks demonstrate that our method performs favorably against state-of-the-art domain adaptive video semantic segmentation methods.

Segment beyond View: Handling Partially Missing Modality for Audio-Visual Semantic Segmentation

Augmented Reality (AR) devices, emerging as prominent mobile interaction platforms, face challenges in user safety, particularly concerning oncoming vehicles. While some solutions leverage onboard camera arrays, these cameras often have limited field-of-view (FoV) with front or downward perspectives. Addressing this, we propose a new out-of-view semantic segmentation task and Segment Beyond View (SBV), a novel audio-visual semantic segmentation method. SBV supplements the visual modality, which miss the information beyond FoV, with the auditory information using a teacher-student distillation model (Omni2Ego). The model consists of a vision teacher utilising panoramic information, an auditory teacher with 8-channel audio, and an audio-visual student that takes views with limited FoV and binaural audio as input and produce semantic segmentation for objects outside FoV. SBV outperforms existing models in comparative evaluations and shows a consistent performance across varying FoV ranges and in monaural audio settings.

Semantic Segmentation in Multiple Adverse Weather Conditions with Domain Knowledge Retention

Semantic segmentation's performance is often compromised when applied to unlabeled adverse weather conditions. Unsupervised domain adaptation is a potential approach to enhancing the model's adaptability and robustness to adverse weather. However, existing methods encounter difficulties when sequentially adapting the model to multiple unlabeled adverse weather conditions. They struggle to acquire new knowledge while also retaining previously learned knowledge. To address these problems, we propose a semantic segmentation method for multiple adverse weather conditions that incorporates adaptive knowledge acquisition, pseudo-label blending, and weather composition replay. Our adaptive knowledge acquisition enables the model to avoid learning from extreme images that could potentially cause the model to forget. In our approach of blending pseudo-labels, we not only utilize the current model but also integrate the previously learned model into the ongoing learning process. This collaboration between the current teacher and the previous model enhances the robustness of the pseudo-labels for the current target. Our weather composition replay mechanism allows the model to continuously refine its previously learned weather information while simultaneously learning from the new target domain. Our method consistently outperforms the state-of-the-art methods, and obtains the best performance with averaged mIoU (%) of 65.7 and the lowest forgetting (%) of 3.6 against 60.1 and 11.3, on the ACDC datsets for a four-target continual multi-target domain adaptation.

Dynamic clustering transformer network for point cloud segmentation

Point cloud segmentation is one of the most important tasks in LiDAR remote sensing with widespread scientific, industrial, and commercial applications. The research thereof has resulted in many breakthroughs in 3D object and scene understanding. Existing methods typically utilize hierarchical architectures for feature representation. However, the commonly used sampling and grouping methods in hierarchical networks are not only time-consuming but also limited to point-wise 3D coordinates, ignoring the local semantic homogeneity of point clusters. To address these issues, we propose a novel 3D point cloud representation network, called Dynamic Clustering Transformer Network (DCTNet). It has an encoder–decoder architecture, allowing for both local and global feature learning. Specifically, the encoder consists of a series of dynamic clustering-based Local Feature Aggregating (LFA) blocks and Transformer-based Global Feature Learning (GFL) blocks. In the LFA block, we propose novel semantic feature-based dynamic sampling and clustering methods, which enable the model to be aware of local semantic homogeneity for local feature aggregation. Furthermore, instead of traditional interpolation approaches, we propose a new semantic feature-guided upsampling method in the decoder for dense prediction. To our knowledge, DCTNet is the first work to introduce semantic information-based dynamic clustering into 3D Transformers. Extensive experiments on an object-based dataset (ShapeNet), and an airborne multispectral LiDAR dataset demonstrate the State-of-the-Art (SOTA) segmentation performance of DCTNet in terms of both accuracy and efficiency. Our code will be made publicly available.

Semantic network analysis of spatial gene sequence in Dabaodao neighbourhood

ABSTRACT Dabaodao neighbourhood was a central gathering place of Chinese in the period of German colonial rule in Qingdao (north-eastern China), so the style and form of it express the integration of Chinese and Western cultures. With the help of the semantic network method, this paper proposes a gene analysis method suitable for historic neighbourhoods and analyses the genetic heritability of the Dabaodao neighbourhood. Regarding specific content, the spatial genes of the Dabaodao neighbourhood are identified as the primary research object, and two genetic research samples are proposed. Next, the spatial texture of the neighbourhoods is analysed to construct a semantic network model. Finally, with the help of the topological quantification method, a specific gene sequence analysis of the Dabaodao neighbourhood is conducted to discuss the degree of genetic influence expressed by it. On this basis, the effectiveness of gene analysis in neighbourhood renewal is discussed in the context of the renewal in the Dabaodao neighbourhood. Semantic network gene analysis not only provides an indicative strategy for the renewal of the Dabaodao neighbourhood but is also a formal analysis method different from the traditional quantitative analysis.

A Unified Visual and Linguistic Semantics Method for Enhanced Image Captioning

Image captioning, also recognized as the challenge of transforming visual data into coherent natural language descriptions, has persisted as a complex problem. Traditional approaches often suffer from semantic gaps, wherein the generated textual descriptions lack depth, context, or the nuanced relationships contained within the images. In an effort to overcome these limitations, we introduce a novel encoder–decoder framework called A Unified Visual and Linguistic Semantics Method. Our method comprises three key components: an encoder, a mapping network, and a decoder. The encoder employs a fusion of CLIP (Contrastive Language–Image Pre-training) and SegmentCLIP to process and extract salient image features. SegmentCLIP builds upon CLIP’s foundational architecture by employing a clustering mechanism, thereby enhancing the semantic relationships between textual and visual elements in the image. The extracted features are then transformed by a mapping network into a fixed-length prefix. A GPT-2-based decoder subsequently generates a corresponding Chinese language description for the image. This framework aims to harmonize feature extraction and semantic enrichment, thereby producing more contextually accurate and comprehensive image descriptions. Our quantitative assessment reveals that our model exhibits notable enhancements across the intricate AIC-ICC, Flickr8k-CN, and COCO-CN datasets, evidenced by a 2% improvement in BLEU@4 and a 10% uplift in CIDEr scores. Additionally, it demonstrates acceptable efficiency in terms of simplicity, speed, and reduction in computational burden.

On the problem of word polysemy in the Khakass language (on the example of the lexicon of expressive verbs)

The subject of the study are polysemantic Khakass verbs of colloquial style, the functional-semantic features of this category of verbs, as well as their semantic modifications, on the basis of which their figurative lexical-semantic variants are formed. Polysemous colloquial verbs, due to their wide situational-pragmatic potential, usually have the most complex, generalising meaning, bordering on desemantisation, and therefore are not always recorded in lexicographic sources. The relevance of the study is due to the fact that there are practically no works devoted to the study of the colloquial and everyday style of the Khakass language. Moreover, this topic is closely related to the concept of expressiveness, which has not been sufficiently studied with the help of Khakass material. The aim of the article is to identify and describe polysemous verbs of the colloquial style from a functional-semantic point of view. In achieving this goal, the following interacting issues were raised: expressive style of language, the phenomenon of polysemy in language, semantic potential of a word, representation of expressive vocabulary in lexicographic sources, good knowledge of the native language as a guarantee of figurative and intuitive perception of the world around us, etc. Methods of distributional analysis and component analysis have been used in the work, which make it possible to identify combination possibilities and semantic features of the described polysemantic verbs, in particular to determine the grammatical, syntactic and semantic methods of forming expressions. In addition, verbs with an expressive frequency not recorded in the Khakass-Russian Dictionary [2006] were identified, as well as lexical-semantic variants of verbs used in everyday speech not listed in the dictionary entries. The material for the analysis were examples collected from the works of Khakass authors. The verbs analysed are, as a rule, a specific part of the colloquial language of the literary characters. Theoretical significance: new knowledge has been obtained in terms of describing polysemantic verbal lexemes of colloquial style, which will serve as theoretical support for further development of the phenomena of polysemy and expressiveness in Khakass and other Turkic languages. Practical significance can be expressed in the use of research materials in the compilation of ideographic dictionaries, textbooks, as well as in teaching activities. In our article, the concepts of "polysemy" and "expressiveness" are discussed within the framework of a summarised understanding of a whole complex of stylistic, emotional, colloquial and other phenomena of language. In this way, an attempt is made to identify the problem for further in-depth and thorough studies of this rich and unique layer of the Khakass lexical system.

A new framework for improving semantic segmentation in aerial imagery

High spatial resolution (HSR) remote sensing imagery presents a rich tapestry of foreground-background intricacies, rendering semantic segmentation in aerial contexts a formidable and vital undertaking. At its core, this challenge revolves around two pivotal questions: 1) Mitigating Background Interference and Enhancing Foreground Clarity. 2) Accurate Segmentation in Dense Small Object Cluster. Conventional semantic segmentation methods primarily cater to the segmentation of large-scale objects in natural scenes, yet they often falter when confronted with aerial imagery’s characteristic traits such as vast background areas, diminutive foreground objects, and densely clustered targets. In response, we propose a novel semantic segmentation framework tailored to overcome these obstacles. To address the first challenge, we leverage PointFlow modules in tandem with the Foreground-Scene (F-S) module. PointFlow modules act as a barrier against extraneous background information, while the F-S module fosters a symbiotic relationship between the scene and foreground, enhancing clarity. For the second challenge, we adopt a dual-branch structure termed disentangled learning, comprising Foreground Precedence Estimation and Small Object Edge Alignment (SOEA). Our foreground saliency guided loss optimally directs the training process by prioritizing foreground examples and challenging background instances. Extensive experimentation on the iSAID and Vaihingen datasets validates the efficacy of our approach. Not only does our method surpass prevailing generic semantic segmentation techniques, but it also outperforms state-of-the-art remote sensing segmentation methods.