Search

In addition to providing critical knowledge of the accurate mass of ions, ion mobility‐mass spectrometry (IM‐MS) delivers complementary data relating to the conformation and size of ions in the form of an ion mobility spectrum and derived parameters, namely, the ion's mobility (K) and the IM‐derived collision cross section (CCS). However, the maximum amount of information obtained in IM‐MS measurements is not currently transferred into analytical databases including the full mobility spectra (CCS distributions) as well as capturing of additional ion species (e.g., adducts) into the same compound entry. We introduce CCSfind, a new tool for building comprehensive databases from experimental IM‐MS measurements of small molecules. CCSfind allows predicted ion species to be chosen for input chemical formulae, which are then targeted by CCSfind after parsing open source mzML input files to provide a unified set of results within a single data processing step. CCSfind can handle both chromatographically separated isomers and IM separation of isomeric ions (e.g., "protomers" or conformers of the same ion species) with simple user control over the output for new database entries in SQL format. Files of up to 1 GB can be processed in less than 2 min on a desktop computer with 32 GB RAM with computational time scaling linearly with the size of the input mzML file or the number of input molecular formulae. Results are manually reviewed, annotated with experimental settings, before committing the database where the full dataset can be retrieved. [ABSTRACT FROM AUTHOR]

With the rapid convergence of the Internet of Things (IoT) and command and control communication networks, there is a swift surge in computation-intensive applications and heterogenous devices. Building an edge computing network near the protocol gateway not only facilitates network protocol conversion, but also alleviates computational load on the gateway. Nevertheless, traditional protocol gateways necessitate the inclusion of conversion methods for all communication protocols in the network for data interoperability with other devices. Consequently, constructing low-complexity protocol conversion programs, and the more efficient and equitable allocation of edge server resources, are critical challenges that need to be addressed. This paper investigates the joint problem of task offloading and heterogeneous protocol device access in command and control communication networks, deriving the benefit parameters of task offloading. We propose a real-time data unified access platform, which, by distributing configuration files, enables the IoT gateway to generate protocol conversion programs, thereby transforming all device protocols into a unified common protocol. Furthermore, to enhance the efficiency of heterogeneous device access to the network, we design a data parsing/encapsulation algorithm based on the real-time data unified access platform, supporting bidirectional mapping of two types of data. Experimental validation confirms the reasonability and effectiveness of the proposed scheme. [ABSTRACT FROM AUTHOR]

Occluded person re‐identification (Re‐ID) is to identify a particular person when the person's body parts are occluded. However, challenges remain in enhancing effective information representation and suppressing background clutter when considering occlusion scenes. This paper proposes a novel attention map‐driven network (AMD‐Net) for occluded person Re‐ID. In AMD‐Net, human parsing labels are introduced to supervise the generation of partial attention maps, while a spatial‐frequency interaction module is suggested to complement the higher‐order semantic information from the frequency domain. Furthermore, a Taylor‐inspired feature filter for mitigating background disturbance and extracting fine‐grained features is proposed. Moreover, a part‐soft triplet loss, which is robust to non‐discriminative body partial features is also designed. Experimental results on Occluded‐Duke, Occluded‐Reid, Market‐1501, and Duke‐MTMC datasets show that this method outperforms existing state‐of‐the‐art methods. The code is available at: https://github.com/ISCLab‐Bistu/SA‐ReID. [ABSTRACT FROM AUTHOR]

Question generation (QG) aims to generate natural questions from relevant input. Existing state-of-the-art QG approaches primarily leverage pre-trained language models (PLMs) to encode the deep semantics within the input. Meanwhile, studies show that the input's dependency parse tree (referred to as syntactic information) is promising in improving NLP-oriented tasks. However, how to incorporate syntactic information in PLMs to guide a QG process effectively still needs to be settled. This paper introduces a syntax-guided sentence-level QG model based on prompt learning. Specifically, we model the syntactic information by utilizing soft prompt learning, jointly considering the syntactic information from a constructed dependency parse graph and PLM to guide question generation. We conduct experiments on two benchmark datasets, SQuAD1.1 and MS MARCO. Experiment results show that our model exceeded both automatic and human evaluation metrics compared with mainstream approaches. Moreover, our case study shows that the model can generate more fluent questions with richer information. [ABSTRACT FROM AUTHOR]

In the modern world, structured and semi-structured knowledge bases hold a considerable amount of data. There-fore, people who are familiar with formal query languages should not be the only ones who can efficiently and clearly query them. Semantic Parsing (SP) is converting natural language utterances into formal meaning representations. The paper suggests a model for SP that uses a novel method of utilizing the Semi-Supervised Generative Adversarial Network (SS-GAN) to enhance the classifier performance. The proposed SS-GAN extends the fine-tuning of word embedding architectures using unlabeled examples in a generative adversarial environment. We provide a regularization strategy for addressing the mode missing problem and unstable training in SS-GAN. The main viewpoint is to use the extracted feature vectors from the discriminator. Hence, the generator produces outputs by aiding the discriminator's learned features. A reconstruction loss is added to the loss function of the SS-GAN to drive the genera-tor to reconstruct outputs from the discriminator's features, hence steering the generator toward actual data configurations. The proposed reconstruction loss improves the performance of SS-GAN, produces high-quality outputs, and may be combined with other regularization loss functions to improve the performance of diverse GANs. We employ BERT word embedding for our model, which can be included in a downstream task and fine-tuned as a model, while the pre-trained BERT model can capture various linguistic properties. We examine the suggested model using the WikiSQL and SparC datasets, and the analysis findings reveal our model outperforms its rivals. The findings from our experiments indicate that the need for labeled samples can be minimized, down to as few as 100 instances, while still achieving commendable classification outcomes. [ABSTRACT FROM AUTHOR]

Unmanned swarm systems are now employed in a wide range of applications. Agriculture, commerce, and the military will all benefit from the advancement of unmanned swarm technology. Scenarios should have a standardized description and definition as system input to support numerous applications and accelerate scenario validation, allowing for more precise and specialized processing to standardize the validation process. A scenario description language for unmanned swarm (SL4U) is introduced in order to implement the unmanned swarm scenario. The scenario is divided into environment and task. The environment is defined, including the task space, target, and home locations, and the initial settings of the unmanned swarm. In terms of tasks, four categories of tasks, four temporal relationship constraints, the event handler, as well as other elements are identified. The dual engines of parsing and micro-embedded collaboration we developed will handle the SL4U. We employ a search and rescue scenario in both simulation and physics to test SL4U. Through the processing of the dual engines, the unmanned swarm's search and rescue operations in a predefined environment are accomplished successfully. In more complex scenarios, just add more descriptions to SL4U and implement them in dual engines rather than starting from scratch. The use of SL4U in verification minimizes the cost and time consumption of unmanned swarm, particularly in physical verification. [ABSTRACT FROM AUTHOR]

Geographic spaces is a vast and complex system involving multiple elements and nonlinear interactions of these elements, and rich in geographical phenomena, processes and patterns. Artificial intelligence methods (AI) are increasingly utilized to extract information of interest, patterns and insights from massive remote sensing (RS) images. Among them, two representative paradigms for RS interpretation are knowledge-driven symbolism and data-driven connectionism. Knowledge-driven approaches are certain and theoretically sound, which consider the regularities, formulae with accompanying theorems, and expert knowledge. However, it is difficult to address huge-scale RS data and complex nonlinear problems. The data-driven paradigm has stronger data learning and representation ability, yet poor interpretability, low trustworthiness of interpretation results, and reliance on massive labeled data. To address these limitations, we argue that current RS intelligent interpretation requires the guidance of geographic ideas and a unified theoretical framework rather than the independent use of remotely sensed data, duplicative development of interpretation models, or isolated applications of geographic information extraction. Thus, we introduce the concepts of trustworthy RS interpretation (TRSI) with multi-granularity of space, time, and attribute to facilitate the understanding of the morphology, types, and indicators (visual perception) and the perspective of the structure, evolution, and trends (geographical cognition) of complex geographical space. Second, integrated technologies of the TRSI with multiple scales and stages are developed, including a pixel-level visual perception, a geo-parcel-level quantitative inversion, and a scene-level geographical cognition. Finally, a RS big data interpretation system with quantitative semantic parsing is designed to support precise applications with cloud–edge-end collaboration. Rather than a simple overview and accumulation of traditional RS intelligent interpretation methods, this paper analyzes current challenges in RS interpretation and proposes new concepts, technologies, and applications of TRSI. It will empower us to perceive the complex geographical space comprehensibly and reveal its structural composition and evolutionary mechanism, serving geospatial decision applications such as evaluation, planning, and forecasting. [ABSTRACT FROM AUTHOR]

Identification of stance has recently gained a lot of attention with the extreme growth of fake news and filter bubbles. Over the last decade, many feature-based and deep-learning approaches have been proposed to solve stance detection. However, almost none of the existing works focus on providing a meaningful explanation for their prediction. In this work, we study stance detection with an emphasis on generating explanations for the predicted stance by capturing the pivotal argumentative structure embedded in a document. We propose to build a stance tree that utilizes rhetorical parsing to construct an evidence tree and to use Dempster Shafer Theory to aggregate the evidence. Human studies show that our unsupervised technique of generating stance explanations outperforms the SOTA extractive summarization method in terms of informativeness, non-redundancy, coverage, and overall quality. Furthermore, experiments show that our explanation-based stance prediction excels or matches the performance of the SOTA model on various benchmark datasets. [ABSTRACT FROM AUTHOR]

This paper introduces an effective method for improving dependency parsing which is based on a graph embedding model. The model helps extract local and global connectivity patterns between tokens. This method allows neural network models to perform better on dependency parsing benchmarks. We propose to incorporate node embeddings trained by a graph embedding algorithm into a bidirectional recurrent neural network scheme. The new model outperforms a baseline reference using a state-of-the-art method on three dependency treebanks for both low-resource and high-resource natural languages, namely Indonesian, Vietnamese and English. We also show that the popular pretraining technique of BERT would not pick up on the same kind of signal as graph embeddings. The new parser together with all trained models is made available under an open-source license, facilitating community engagement and advancement of natural language processing research for two low-resource languages with around 300 million users worldwide in total. [ABSTRACT FROM AUTHOR]

Image animation creates visually compelling effects by animating still source images according to driving videos. Recent work performs animation on arbitrary objects using unsupervised methods and can relatively robustly perform motion transfer on human bodies. However, the complex representation of motion and unknown correspondence between human bodies often lead to issues such as distorted limbs and missing semantics, which make human animation challenging. In this paper, we propose a semantically guided, unsupervised method of motion transfer, which uses semantic information to model motion and appearance. Specifically, we use a pre-trained human parsing network to encode the rich and diverse foreground semantic information, thus generating fine details. Secondly, we use a cross-modal attention layer to learn the semantic region's correspondence between human bodies to guide the network in selecting appropriate input features, prompting the network to generate accurate results. Experiments demonstrate that our method outperforms state-of-the-art methods in motion-related metrics, while effectively addressing the problems of semantic missing and unclear limb structures prevalent in human motion transfer. These improvements can facilitate its applications in various fields, such as education and entertainment. [Display omitted] • Proposes a novel framework for human image animation using semantic features and cross-modal attention. • Introduces semantic segmentation features to represent motion and identity information. • Employs a cross-modal attention mechanism to establish correspondences between semantic regions. • Achieves state-of-the-art performance on complex human motions like TaiChi poses. • Reduces issues of missing semantics and distorted limbs common in human animation. [ABSTRACT FROM AUTHOR]

Human parsing is very important in a diverse range of industrial applications. Despite the considerable progress that has been achieved, the performance of existing methods is still less than satisfactory, since these methods learn the shared features of various parsing labels at the image level. This limits the representativeness of the learnt features, especially when the distribution of parsing labels is imbalanced or the scale of different labels is substantially different. To address this limitation, a Region‐level Parsing Refiner (RPR) is proposed to enhance parsing performance by the introduction of region‐level parsing learning. Region‐level parsing focuses specifically on small regions of the body, for example, the head. The proposed RPR is an adaptive module that can be integrated with different existing human parsing models to improve their performance. Extensive experiments are conducted on two benchmark datasets, and the results demonstrated the effectiveness of our RPR model in terms of improving the overall parsing performance as well as parsing rare labels. This method was successfully applied to a commercial application for the extraction of human body measurements and has been used in various online shopping platforms for clothing size recommendations. The code and dataset are released at this link https://github.com/applezhouyp/PRP. [ABSTRACT FROM AUTHOR]

It is important to establish solid mathematical knowledge throughout the education period. Matrices are important mathematical objects commonly used in many diverse disciplines, including mathematics, engineering, and science. Most problems encountered in such disciplines are represented by mathematical models with various types of matrices, and solved through some applications of matrix algebra. Although simple or advanced operations of matrices can be performed by using modern programming languages, it usually results in a large fragment of code with a low level of readability due to a complicated sequence of control statements. On the other hand, special-purpose languages handle these operations via library functions, presenting poor integration with other programming environments, and less programming flexibility and practice. This paper addresses the design and development of a programming language, called MxPL, which supports matrix-related mathematics with the provision of some basic structures and functions. Firstly, a grammar that is compatible with the usual notations of matrices is constructed and the parser is produced. Then, the code verifier and interpreter for MxPL programs are implemented. Some code examples are presented to illustrate the performance of several sophisticated matrix operations. The comparative analysis of MxPL is conducted with modern programming languages based on language features. [ABSTRACT FROM AUTHOR]

Background: Knowledge graphs are well-suited for modeling complex, unstructured, and multi-source data and facilitating their analysis. During the COVID-19 pandemic, adverse event data were integrated into a knowledge graph to support vaccine safety surveillance and nimbly respond to urgent health authority questions. Here, we provide details of this post-marketing safety system using public data sources. In addition to challenges with varied data representations, adverse event reporting on the COVID-19 vaccines generated an unprecedented volume of data; an order of magnitude larger than adverse events for all previous vaccines. The Patient Safety Knowledge Graph (PSKG) is a robust data store to accommodate the volume of adverse event data and harmonize primary surveillance data sources. Methods: We designed a semantic model to represent key safety concepts. We built an extract-transform-load (ETL) data pipeline to parse and import primary public data sources; align key elements such as vaccine names; integrated the Medical Dictionary for Regulatory Activities (MedDRA); and applied quality metrics. PSKG is deployed in a Neo4J graph database, and made available via a web interface and Application Programming Interfaces (APIs). Results: We import and align adverse event data and vaccine exposure data from 250 countries on a weekly basis, producing a graph with 4,340,980 nodes and 30,544,475 edges as of July 1, 2022. PSKG is used for ad-hoc analyses and periodic reporting for several widely available COVID-19 vaccines. Analysis code using the knowledge graph is 80% shorter than an equivalent implementation written entirely in Python, and runs over 200 times faster. Conclusions: Organizing safety data into a concise model of nodes, properties, and edge relationships has greatly simplified analysis code by removing complex parsing and transformation algorithms from individual analyses and instead managing these centrally. The adoption of the knowledge graph transformed how the team answers key scientific and medical questions. Whereas previously an analysis would involve aggregating and transforming primary datasets from scratch to answer a specific question, the team can now iterate easily and respond as quickly as requests evolve (e.g., "Produce vaccine-X safety profile for adverse event-Y by country instead of age-range"). [ABSTRACT FROM AUTHOR]

Most cropping-and-segmenting pattern parsers typically establish a single metric/scheme to reason diverse inner correlations, resulting in over-general and redundant representations. To make pattern parsing more streamlined and efficient, a fragile correlation pruner network (FCPN) with correlation-steered attention shifters (CSASs) and graph attention expectation-maximum routing agreement (GAEMRA) is proposed. CSASs prune fragile (weak and substitutable) part-to-whole correlations. They stipulate that only those primary entities (representing components) fulfilling the criteria of inter-part diversity and intra-object cohesiveness can update senior entities (representing the whole/intermediate composites). To further boost effects, GAEMRA is defined to shield the redundant voting signals of conventional routing agreement. With CSASs and GAEMRA, FCPN gradually parses objective semantic patterns by clustering highly associated secondary entities in a bottom-up "part backtracking" manner. Quantitative and ablation experiments surrounding face and human parsing demonstrate the superiority of FCPN over the state-of-the-arts, especially for the definition of fine-grained semantic boundaries. [ABSTRACT FROM AUTHOR]

Automated log parsing is essential for many log-mining applications, as logs provide a vast range of information on events and variations within an operating system or software at runtime. Over the years, various methods have been proposed for log parsing. With improved log-parsing methods, log-mining applications can gain deeper insights into system behaviors and identify anomalies or failures promptly. However, current log parsers still face limitations, such as insufficient parsing of log templates and a lack of parallelism, as well as inaccurate log template parsing. To overcome these limitations, we have designed Polo, a parser that leverages a prefix forest composed of ternary search trees to mine templates from logs. We then conducted extensive experiments to evaluate the accuracy of Polo on nine representative system logs, achieving an average accuracy of 0.987. It is 9.93% to 40.95% faster than the state-of-the-art parsing methods. Furthermore, we evaluated our approach on a downstream log analysis task, specifically anomaly detection. The experimental results demonstrated that, in terms of F1-score, our parser outperformed Deeplog, LogAnomaly, CNN, and LogRobust by 11.5%, 4%, 1%, and 19.1%, respectively, exhibiting a promising recall score of 0.971. These results indicate the effectiveness of Polo for anomaly detection. [ABSTRACT FROM AUTHOR]

One of the main challenges in target-dependent sentiment classification (TDSC) is dealing with sentences that contain multiple targets with varying polarities. Traditional sentiment analysis has shown the effectiveness of language characteristics. Therefore, we propose a method to extract target semantic-related tokens from sentences in order to simplify the sentiment classification task. To achieve this, we establish six grammatical principles that utilize grammatical knowledge to filter the relevant descriptions of targets. Since a target is typically a noun and acts as a subject, we summarize the six rules to extract the contexts contained in the objects and subordinate clauses. We use dependency parsing to analyze the grammatical relations between the target and its context. We design a data pre-processing method called Text Filtering (TF) to automate this procedure. After executing the TF algorithm, we pass the target-related words to a simple classifier to predict their sentiment polarities. Rather than feeding these features directly to a network and letting it learn features on its own, our approach employs dependency relations to extract context linked to the target. This provides the network with meaningful and representative features, resulting in superior results. We conduct ablation studies to investigate the effectiveness of the proposed TF algorithm. In the restaurant hard dataset, our approach improves accuracy by 13.76% and macro-F1 by 14.65% compared to a CNN-based method where TF is not implemented. • Incorporate linguistic knowledge to the model by a two-step framework for effective handling of multiple targets. • Target-related contexts were extracted by the summarized rules which distinguish content and function words. • A hierarchical and recursive data process method was designed specifically for TDSC named Text filtering (TF). [ABSTRACT FROM AUTHOR]

In the era of GeoAI, Geospatial Intelligent Question Answering (GeoIQA) represents the ultimate pursuit for everyone. Even generative AI systems like ChatGPT-4 struggle to handle complex GeoIQA. GeoIQA is domain complex IQA, which aims at understanding and answering questions accurately. The core of IQA is the Question Classification (QC), which mainly contains four types: content-based, template-based, calculation-based and method-based classification. These IQA_QC frameworks, however, struggle to be compatible and integrate with each other, which may be the bottleneck restricting the substantial improvement of IQA performance. To address this problem, this paper reviewed recent advances on IQA with the focus on solving question classification and proposed a comprehensive IQA_QC framework for understanding user query intention more accurately. By introducing the basic idea of the IQA mechanism, a three-level question classification framework consisting of essence, form and implementation is put forward which could cover the complexity and diversity of geographical questions. In addition, the proposed IQA_QC framework revealed that there are still significant deficiencies in the IQA evaluation metrics in the aspect of broader dimensions, which led to low answer performance, functional performance and systematic performance. Through the comparisons, we find that the proposed IQA_QC framework can fully integrate and surpass the existing classification. Although our proposed classification can be further expanded and improved, we firmly believe that this comprehensive IQA_QC framework can effectively help researchers in both semantic parsing and question querying processes. Furthermore, the IQA_QC framework can also provide a systematic question-and-answer pair/library categorization system for AIGCs, such as GPT-4. In conclusion, whether it is explicit GeoAI or implicit GeoAI, the IQA_QC can play a pioneering role in providing question-and-answer types in the future. [ABSTRACT FROM AUTHOR]

The rapid development of information technology has made the amount of information in massive texts far exceed human intuitive cognition, and dependency parsing can effectively deal with information overload. In the background of domain specialization, the migration and application of syntactic treebanks and the speed improvement in syntactic analysis models become the key to the efficiency of syntactic analysis. To realize domain migration of syntactic tree library and improve the speed of text parsing, this paper proposes a novel approach—the Double-Array Trie and Multi-threading (DAT-MT) accelerated graph fusion dependency parsing model. It effectively combines the specialized syntactic features from small-scale professional field corpus with the generalized syntactic features from large-scale news corpus, which improves the accuracy of syntactic relation recognition. Aiming at the problem of high space and time complexity brought by the graph fusion model, the DAT-MT method is proposed. It realizes the rapid mapping of massive Chinese character features to the model's prior parameters and the parallel processing of calculation, thereby improving the parsing speed. The experimental results show that the unlabeled attachment score (UAS) and the labeled attachment score (LAS) of the model are improved by 13.34% and 14.82% compared with the model with only the professional field corpus and improved by 3.14% and 3.40% compared with the model only with news corpus; both indicators are better than DDParser and LTP 4 methods based on deep learning. Additionally, the method in this paper achieves a speedup of about 3.7 times compared to the method with a red-black tree index and a single thread. Efficient and accurate syntactic analysis methods will benefit the real-time processing of massive texts in professional fields, such as multi-dimensional semantic correlation, professional feature extraction, and domain knowledge graph construction. [ABSTRACT FROM AUTHOR]

Relation classification is an important task in information extraction that involves identifying potential semantic connections between two entities in a sentence. Most relation extraction models either use semantic or structure information as the relation representation for classification. Although some researchers have tried to fuse both types of information, they simply combine different information as relation representation, which ignores semantic and structure information contribute unequally to the relation representation at the instance level, leading to degradation of performance and generalization capability. To address this issue, we propose a fusing multi-information model with a self-attentive strategy (Att-FMI) for relation classification. Our approach utilizes the pre-trained BERT model and dependency syntactic parser to obtain semantic and structure representations from the sequence, respectively and efficiently combines the two with support the selfattention mechanism. Experimental results on four commonly used datasets, including TACRED, TACREV, KBP-37 and SemEval-2010 Task8, demonstrate that our approach yields significant improvements over baseline methods. Additionally, we observe that the Att-FMI model exhibits greater robustness against information interference in the extraction of entity relations from longer sentences than previous methods. [ABSTRACT FROM AUTHOR]

The article discusses DomainNet, a method for homograph detection and disambiguation in data lakes, addressing the challenge of identifying data values with multiple meanings in heterogeneous repositories. Topics include leveraging network centrality measures on a bipartite graph to distinguish homographs, proposing a novel approach for determining the number of meanings, and exploring the impact of homograph identification on entity-matching and domain discovery tasks within data lakes.

Domain-specific languages allow engineers and domain experts to express problems and design solutions using domain-focused vocabularies and abstractions, by means of graphical or textual syntaxes. In the case of textual syntaxes, language engineers can opt for creating a language-specific syntax by defining and maintaining a BNF-style grammar, or use an existing general-purpose reflective syntax such as the XML Metadata Interchange (XMI) or the Human Usable Textual Notation (HUTN), which do not require any development and maintenance effort, but which are more verbose and cannot be customised. We present Flexmi: a new general-purpose textual syntax for defining models that conform to Eclipse Modelling Framework's Ecore-based metamodels. Flexmi offers XML and YAML/JSON syntax flavours, it can be fuzzily parsed to reduce verbosity, and it includes a templating system to facilitate encapsulation of reusable composite model element structures, thus enabling more concise model specifications. We have evaluated Flexmi for verbosity and model loading performance against XMI, HUTN, and a bespoke (i.e. custom) textual syntax for Ecore (Emfatic). Our results indicate that the use of fuzzy parsing and templates allow Flexmi to achieve a significant reduction in the verbosity of models compared to XMI/HUTN and can become almost as concise as a bespoke textual syntax, with a moderate performance penalty. [ABSTRACT FROM AUTHOR]

• First investigation on machine reading comprehension framework for Chinese spoken language understanding. • Development of the TCM_CLS module with concise architecture for efficient multi-source information fusion. • Multi-task learning for joint slot filling and intent detection using adversarial training and focal loss. • Achievement of new SOTA performance compared with baseline methods. Spoken language understanding (SLU) plays a crucial role in the performance of dialogue systems. It usually includes slot filling and intent detection (SFID) tasks aiming at semantic parsing of utterances. At present, researchers focus mainly on English SLU tasks, while such investigations on Chinese utterances are not sufficient. In this paper, we first propose a reading comprehension based multi-source information fusion network, called ReCoMIF for Chinese SFID tasks by transforming the SLU task into a multi-turn question answering procedure comprising multiple choice for intent detection and span extraction for slot filling. Moreover, we present a TCM_CLS module with a concise architecture composed of TextCNN, MaxPooling, and feed forward network plus the [CLS] representation. Such three TCM_CLS modules are stacked in the proposed ReCoMIF network that can serve as sufficient integration of multi-source information originating from contexts, reading comprehension based queries, and hidden representations concerning intent and slot semantics. Finally, experimental results and ablation studies on three Chinese SLU datasets show that our proposed model can effectively fuse intent and slot information achieved by\ state-of-the-art performance compared with other baseline methods. [ABSTRACT FROM AUTHOR]

System logs are a crucial component of system maintainability, as they record the status of the system and essential events for troubleshooting and maintenance when necessary. Therefore, anomaly detection of system logs is crucial. Recent research has focused on extracting semantic information from unstructured log messages for log anomaly detection tasks. Since BERT models work well in natural language processing, this paper proposes an approach called CLDTLog, which introduces contrastive learning and dual-objective tasks in a BERT pre-trained model and performs anomaly detection on system logs through a fully connected layer. This approach does not require log parsing and thus can avoid the uncertainty caused by log parsing. We trained the CLDTLog model on two log datasets (HDFS and BGL) and achieved F1 scores of 0.9971 and 0.9999 on the HDFS and BGL datasets, respectively, which performed better than all known methods. In addition, when using only 1% of the BGL dataset as training data, CLDTLog still achieves an F1 score of 0.9993, showing excellent generalization performance with a significant reduction of the training cost. [ABSTRACT FROM AUTHOR]

In this paper, we proposed a new method for image-based grammatical inference of deterministic, context-free L-systems (D0L systems) from a single sequence. This approach is characterized by first parsing an input image into a sequence of symbols and then, using a genetic algorithm, attempting to infer a grammar that can generate this sequence. This technique has been tested using our test suite and compared to similar algorithms, showing promising results, including solving the problem for systems with more rules than in existing approaches. The tests show that it performs better than similar heuristic methods and can handle the same cases as arithmetic algorithms. [ABSTRACT FROM AUTHOR]

In organizational theory, institutionalists generally make predictions of corresponding context and policy outcome based on structural processes. Psychoanalytic theory, in contrast, focuses on the rhetorical framing rather than the environment of a policy for predictive outcomes. This study aims to explore the debate over policy prediction by developing a supervised machine learning model to predict for policy success and context in the United Nations (UN). Through data collected with a python web scraper on all UN meetings in the General Assembly (GA) and Security Council (SC) between 1994 and 2020, we parse motions, policies, and conflict indicators, before passing meeting records through the Linguistic Inquiry and Word Count (LIWC) psycholinguistic algorithm. Next, we build 12 different machine learning models to predict for policy passage and context using preprocessed motion and LIWC data; results demonstrate that the psychoanalytic models better predicted for both context and policy outcomes than the institutionalist models, suggesting that the classical political axiom, "actions speak louder than words," may not be supported by the empirical evidence. [ABSTRACT FROM AUTHOR]

A new parsing method based on the semi-Thue system is described. Similar to, but with more efficient implementation than Markov normal algorithms, it can be used for parsing any recursively enumerable language. Despite its computational power, it is meant to be used primarily for parsing programming and domain-specific languages. It enables a straightforward simulation of a number of existing parsing algorithms based on context-free grammars. The list includes both top-down shift-produce methods (such as SLL and LL) and bottom-up shift-reduce methods (such as LALR and LR), as well as mixed top-down-and-bottom-up methods such as LLLR. To justify the use of the new parsing method, the paper provides numerous examples of how a parser can actually be made in practice. It is advised that the main part of the parser is based on some simple well-established approach, e.g., SLL(1), while syntactically more complicated phrases can be parsed by exploiting the full power of the new parser. These phrases may either be extensions to the original language or some embedded domain-specific language. In all such and similar cases, no part of the language is restricted to be context-free. In fact, context-sensitive languages can be handled quite efficiently. [ABSTRACT FROM AUTHOR]

In computer vision, human parsing is challenging due to its demand for accurate human region location and semantic partitioning. This dense prediction task needs powerful computation and high‐precision models. To enable real‐time parsing on resource‐limited devices, the authors introduced a lightweight model using ResNet18 as a core network. The authors simplified the pyramid module, improving context clarity and reducing complexity. The authors integrated a spatial attention fusion strategy to counter precision loss in the light‐weighting process. Traditional models, despite their segmentation precision, are limited by their computational complexity and extensive parameters. The authors implemented knowledge distillation (KD) techniques to enhance the authors' lightweight network's accuracy. Traditional methods can fail to learn useful knowledge with significant network differences. Hence, the authors used a novel distillation approach based on inter‐class and intra‐class relations in prediction outcomes, noticeably improving parsing accuracy. The authors' experiments on the Look into Person (LIP) dataset show that their lightweight model significantly reduces parameters while maintaining parsing precision and enhancing inference speed. [ABSTRACT FROM AUTHOR]

As an effective method to assist intelligent management of public areas, person re-identification (re-ID) technology has been developed rapidly in recent years. However, there are still some insurmountable obstacles in practical applications, among which misalignment caused by some factors is a challenging problem. Unlike previous approaches that only superficially mine pedestrian information to solve this problem, we propose an aligned person re-ID method based on human semantic parsing and message passing. Our method achieves pixel-level alignment through the incorporation of semantic parsing and also utilizes the results of semantic parsing. It constructs a graph neural network based on the structure of the human body to achieve information interaction between various part features. Additionally, various semantic features and a global feature are considered and used in the loss function for the ensemble of features, thereby ensuring the discrimination and robustness. Such ensemble learning allows our method to perform well not only for the unaligned case, but also have the ability to handle occlusion. Thus, the proposed SPMP method achieves better performance than most existing methods on multiple popular datasets. [ABSTRACT FROM AUTHOR]

In this paper, we describe a model for the automatic generation of literary sentences in French. Although there has been much recent effort directed towards automatic text generation in general, the generation of creative, literary sentences that is not restricted to a specific genre, which we approached in this work, is a difficult task that is not commonly treated in the scientific literature. In particular, our present model has not been previously applied to the generation of sentences in the French language. Our model was based on algorithms that we previously used to generate sentences in Spanish and Portuguese and on a new corpus, which we constructed and present here, consisting of literary texts in French, called MegaLitefr. Our automatic text generation algorithm combines language models, shallow parsing, the canned text method, and deep learning artificial neural networks. We also present a manual evaluation protocol that we propose and implemented to assess the quality of the artificial sentences generated by our algorithm, by testing if they fulfil four simple criteria. We obtained encouraging results from the evaluators for most of the desired features of our artificially generated sentences. [ABSTRACT FROM AUTHOR]

• Sequence-to-sequence constituent parsers are lagging behind task-specific techniques. • We notably improve their accuracy by introducing a powerful neural architecture. • For the first time, we extend their coverage to deal with discontinuous constituent parsing. • Our text-to-parse transducer achieves a competitive performance on the main benchmarks. In order to achieve deep natural language understanding, syntactic constituent parsing is a vital step, highly demanded by many artificial intelligence systems to process both text and speech. One of the most recent proposals is the use of standard sequence-to-sequence models to perform constituent parsing as a machine translation task, instead of applying task-specific parsers. While they show a competitive performance, these text-to-parse transducers are still lagging behind classic techniques in terms of accuracy, coverage and speed. To close the gap, we here extend the framework of sequence-to-sequence models for constituent parsing, not only by providing a more powerful neural architecture for improving their performance, but also by enlarging their coverage to handle the most complex syntactic phenomena: discontinuous structures. To that end, we design several novel linearizations that can fully produce discontinuities and, for the first time, we test a sequence-to-sequence model on the main discontinuous benchmarks, obtaining competitive results on par with task-specific discontinuous constituent parsers and achieving state-of-the-art scores on the (discontinuous) English Penn Treebank. [ABSTRACT FROM AUTHOR]

Dependency parsing is a crucial step towards deep language understanding and, therefore, widely demanded by numerous Natural Language Processing applications. In particular, left-to-right and top-down transition-based algorithms that rely on Pointer Networks are among the most accurate approaches for performing dependency parsing. Additionally, it has been observed for the top-down algorithm that Pointer Networks' sequential decoding can be improved by implementing a hierarchical variant, more adequate to model dependency structures. Considering all this, we develop a bottom-up oriented Hierarchical Pointer Network for the left-to-right parser and propose two novel transition-based alternatives: an approach that parses a sentence in right-to-left order and a variant that does so from the outside in. We empirically test the proposed neural architecture with the different algorithms on a wide variety of languages, outperforming the original approach in practically all of them and setting new state-of-the-art results on the English and Chinese Penn Treebanks for non-contextualized and BERT-based embeddings. • The left-to-right transition-based algorithm is a state-of-the-art dependency parser. • This is based on Pointer Networks that implement a sequential decoding. • We develop a hierarchical decoding for the left-to-right transition-based parser. • This approach reduces error-propagation by providing long-range dependency information. • The resulting neural architecture achieves a state-of-the-art performance on Penn Treebanks. [ABSTRACT FROM AUTHOR]

A machine learning model for correcting errors in Ukrainian texts has been developed. It was established that the neural network has the ability to correct simple sentences written in Ukrainian; however, the development of a full-fledged system requires the use of spell-checking using dictionaries and the checking of rules, both simple and those based on the result of parsing dependencies or other features. In order to save computing resources, a pre-trained BERT (Bidirectional Encoder Representations from Transformer) type neural network was used. Such neural networks have half as many parameters as other pre-trained models and show satisfactory results in correcting grammatical and stylistic errors. Among the ready-made neural network models, the pre-trained neural network model mT5 (a multilingual variant of T5 or Text-to-Text Transfer Transformer) showed the best performance according to the BLEU (bilingual evaluation understudy) and METEOR (metric for evaluation of translation with explicit ordering) metrics. [ABSTRACT FROM AUTHOR]

Recently, open-domain question-answering systems have achieved tremendous progress because of developments in large language models (LLMs), and have successfully been applied to question-answering (QA) systems, or Chatbots. However, there has been little progress in open-domain question answering in the geographic domain. Existing open-domain question-answering research in the geographic domain relies heavily on rule-based semantic parsing approaches using few data. To develop intelligent GeoQA agents, it is crucial to build QA systems upon datasets that reflect the real users' needs regarding the geographic domain. Existing studies have analyzed geographic questions using the geographic question corpora Microsoft MAchine Reading Comprehension (MS MARCO), comprising real-world user queries from Bing in terms of structural similarity, which does not discover the users' interests. Therefore, we aimed to analyze location-related questions in MS MARCO based on semantic similarity, group similar questions into a cluster, and utilize the results to discover the users' interests in the geographic domain. Using a sentence-embedding-based topic modeling approach to cluster semantically similar questions, we successfully obtained topic models that could gather semantically similar documents into a single cluster. Furthermore, we successfully discovered latent topics within a large collection of questions to guide practical GeoQA systems on relevant questions. [ABSTRACT FROM AUTHOR]

Event extraction is an essential task in natural language processing. Although extensively studied, existing work shares issues in three aspects, including (1) the limitations of using original syntactic dependency structure, (2) insufficient consideration of the node level and type information in Graph Attention Network (GAT), and (3) insufficient joint exploitation of the node dependency type and part-of-speech (POS) encoding on the graph structure. To address these issues, we propose a novel framework for open event extraction in documents. Specifically, to obtain an enhanced dependency structure with powerful encoding ability, our model is capable of handling an enriched parallel structure with connected ellipsis nodes. Moreover, through a bidirectional dependency parsing graph, it considers the sequence of order structure and associates the ancestor and descendant nodes. Subsequently, we further exploit node information, such as the node level and type, to strengthen the aggregation of node features in our GAT. Finally, based on the coordination of triple-channel features (i.e., semantic, syntactic dependency and POS), the performance of event extraction is significantly improved. Extensive experiments are conducted to validate the effectiveness of our method, and the results confirm its superiority over the state-of-the-art baselines. Furthermore, in-depth analyses are provided to explore the essential factors determining the extraction performance. [ABSTRACT FROM AUTHOR]

Attribute-level sentiment element extraction aims to obtain the word pair < opinion target, opinion word > from texts, which mainly obtain fine-grained evaluation information in the attribute level. Due to the information fragmentation and semantic sparseness of product reviews, it is difficult to capture more comprehensive local information from unstructured texts, which leads to the incorrect extraction of some word pairs. Aimed at the problem, this paper proposes a framework for Attribute-Level Sentiment Element Extraction (ALSEE) towards product reviews. Firstly, a small amount of sample data is selected by random sampling, and multiple features (including part of speech, word distance, dependency relationship and semantic role) which are labelled. The labelled data are used as the training set. Then, the Condition Random Field (CRF) model is applied to extract opinion targets (OT) and opinion words (OW). The self-training strategy is used to achieve the semi-supervised learning of CRF model through iterative training. Finally, target-opinion word pairs with modifying relationship are obtained by dependency parsing. Compared with the existing methods, the proposed framework can effectively extract attribute-level sentiment elements though experimental results. [ABSTRACT FROM AUTHOR]

Abstractions are significant domain terms that have assisted in requirements elicitation and modeling. To extend the assistance toward requirements validation, we present in this paper an automated approach to identifying the abstractions for supporting requirements-based testing. We select relevant Wikipedia pages to serve as a domain corpus that is independent from any specific software system. We further define five novel patterns based on part-of-speech tagging and dependency parsing, and frame our candidate abstractions in the form of pairs for better testability, where the "key" helps locate "what to test", and the "value" helps guide "how to test it" by feeding in concrete data. We evaluate our approach with six software systems in two application domains: Electronic health records and Web conferencing. The results show that our abstractions are more accurate than those generated by a state-of-the-art technique. While the initial findings indicate our abstractions' capabilities of revealing bugs and matching the environmental assumptions created manually, we articulate a new way to perform requirements-based testing by focusing on a software system's changing features. Specifically, we hypothesize that the same feature would behave differently under a pair of opposing environmental conditions and assess our abstractions' applicability to this new form of feature testing. [ABSTRACT FROM AUTHOR]

To fully pay attention to identity-sensitive feature information and utilize the correlations of inter-attributes and attributes-body parts, this paper proposes a person re-identification (re-ID) method based on the construction of graph convolutional network (GCN) with crucial attribute feature and body parts. First, it establishes the multiscale context-aware network (MSCAN) using dilated convolution with different expansion ratios, which can learn multiscale context information and obtain diversified global features. Subsequently, the human parsing model is utilized to extract the body part features. According to the attribute importance degree, the paper constructs low-dimensional GCN integrating the vital attributes and body parts of person descriptions to obtain discriminative local features. Finally, based on attribute prediction, it reduces the range of the images to be matched with discriminating possible objects from query images, thereby simplifying retrieval process. The experimental results demonstrate that the novel designed method can effectively improve person re-ID performance and achieve competitive evaluation results on typical public testing datasets. [ABSTRACT FROM AUTHOR]

We present a one-stage F ully C onvolutional Li ne P arsing network (F-Clip) that detects line segments from images. The proposed network is very simple and flexible with variations that gracefully trade off between speed and accuracy for different applications. F-Clip detects line segments in an end-to-end fashion by predicting each line's center position, length, and angle. We further customize the design of convolution kernels of our fully convolutional network to effectively exploit the statistical priors of the distribution of line angles in real image datasets. We conduct extensive experiments and show that our method achieves a significantly better trade-off between efficiency and accuracy, resulting in a real-time line detector at up to 73 FPS on a single GPU. Such inference speed makes our method readily applicable to real-time tasks without compromising any accuracy of previous methods. Moreover, when equipped with a performance-improving backbone network, F-Clip is able to significantly outperform all state-of-the-art line detectors on accuracy at a similar or even higher frame rate. In other word, under same inference speed, F-Clip always achieving best accuracy compare with other methods. Source code https://github.com/Delay-Xili/F-Clip. [ABSTRACT FROM AUTHOR]

Unlike other mildly context-sensitive formalisms, Combinatory Categorial Grammar (CCG) cannot be parsed in polynomial time when the size of the grammar is taken into account. Refining this result, we show that the parsing complexity of CCG is exponential only in the maximum degree of composition. When that degree is fixed, parsing can be carried out in polynomial time. Our finding is interesting from a linguistic perspective because a bounded degree of composition has been suggested as a universal constraint on natural language grammar. Moreover, ours is the first complexity result for a version of CCG that includes substitution rules, which are used in practical grammars but have been ignored in theoretical work. [ABSTRACT FROM AUTHOR]

Weeds have been one of the main factors to affect the growth of crops in the seedling stage. Timely weeding is a necessary measure to ensure crop yield. An intelligent field weeding equipment can also be a promising potential deployment in the unmanned farm system at the current stage of intelligent agriculture. Effective recognition of crops and weeds has been a high demand to promote the development of intelligent weeding equipment. Previous research was focused mainly on object detection and semantic segmentation using deep learning. A great challenge is still remained in the performance of target detection, in the case of overlap images between the crops and weeds under the complex field. The reason was that the different target areas cannot be further divided when the generated anchor box overlaps in a large area. The pixel level annotation can also be required to train the semantic segmentation, where the data samples cannot be easy to obtain. The weak real-time performance cannot be conducive to practical application. In this study, an improved model was proposed using shifted window Transformer (Swin Transformer) network, in order to enhance the accuracy and real-time performance of crop and weed recognition. The specific procedure was as follows. 1) A semantic segmentation model of corn was established for the real and complex field scene. The backbone of the model was the Swin Transformer architecture, which was denoted by Swin-Base. The full self-attention mechanism was also adopted to significantly enhance the modeling ability in the Swin Transformer using the shift window division configuration. Self-attention was then calculated locally in the non-overlapping window of the segmented image block, where the cross-window connection was allowed. The computational complexity of the backbone presented a linear relationship with the image size, thereby elevating the inference speed of the model. The hierarchical feature representation was constructed through the Swin Transformer for the dense prediction of the model at the pixel level. 2) The Unified perceptual parsing Network (UperNet) was used as an efficient semantic segmentation framework. Among them, the feature extractor was the Feature Pyramid Network (FPN) using the Swin Transformer backbone. The multi-level features obtained by Swin Transformer were used by the FPN to represent the corresponding pyramid level. An effective global prior feature expression was added in the Pyramid Pooling Module (PPM). Better performance of semantic segmentation was achieved using the fusion of the hierarchical semantic information. The Swing transformer backbone and UperNet framework were combined into one model through the Decoder-Encoder structure, denoted by Swin-Base-UN. 3) The structure of the Swin-Base backbone was improved to enhance the inference speed. The number of network parameters and calculation cost were reduced to decrease the number of hidden layer channels, headers, and Swin Transformer blocks. Therefore, four improved models were generated, including the Swin-Large-UN, Swin-Small-UN, Swin-Tiny-UN, and Swin-Nano-UN. The model size and computational complexity of improved models were about 2, 1/2, 1/4, and 1/8 times of Swin-Base-UN, respectively. 4) Taking the segmentation of corn morphological region as the case study, an improved image morphological processing combination was established to recognize and segment all the weed regions in real time. The segmentation of corn was also used to segment the weeds. The weed pixel annotation was removed from the training data of the model. As such, a large number of annotation data at the pixel level was obtained in the semantic segmentation of the improved model, compared with the original one. A comparison was made on the performance of all models in training, validation, and testing. Consequently, the Swin-Tiny-UN was determined as the best model to achieve the optimal balance between accuracy and speed. Specifically, the mean Intersection over Union (mIoU) and mean Pixel Accuracy (mPA) on the test set were 94.83% and 97.18%, respectively, which increased by 3.27 and 4.71 percentage points, respectively, compared with the RestNet-101-UN using traditional Convolutional Neural Networks (CNN) backbone. The inference speed of the model was achieved by 18.94 frames/s. The best model of semantic segmentation was superior to the traditional one, in terms of the region segmentation accuracy, pixel recognition accuracy, and inference speed. The image segmentation showed that the improved model can be expected to accurately recognize and segment maize and weeds in complex field scenes. The average correct detection rate of the improved model was 95.04% for the video stream data in the process of field work, whereas, the average detection time per frame was 5.51×10-2 s. Consequently, the improved model can be expected to detect the corn and weeds in the process of field work, indicating higher accuracy and real-time performance under practical application conditions. The findings can provide a strong reference for the development of intelligent weeding equipment. [ABSTRACT FROM AUTHOR]

The limitations of traditional parsing architecture are well known. Even when paired with parsing methods that accept all context-free grammars (CFGs), the resulting combination for any given CFG accepts only a limited subset of corresponding character-level context-free languages (CFL). We present a novel scanner-based architecture that for any given CFG accepts all corresponding character-level CFLs. It can directly parse all possible specifications consisting of a grammar and regular definitions. The architecture is based on right-nulled generalized LR (RNGLR) parsing and is a generalization of the context-aware scanning architecture. Our architecture does not require any disambiguation rules to resolve lexical conflicts, it conceptually has an unbounded parser and scanner lookahead and it is streaming. The added robustness and flexibility allow for easier grammar development and modification. [ABSTRACT FROM AUTHOR]

There are few studies on automated parsing of Chinese privacy policies, and the extracted information is not comprehensive enough which cannot meet the demand of management for automated review of Android App privacy policies. In order to realize the automated parsing of Chinese privacy policies, this paper proposed a Chinese privacy policy parsing algorithm based on dependent syntax, called PPExtractor. Based on two types of seed patterns, verb and preposition, PPExtractor used the bootstrap pattern method to generate the set of privacy event patterns. Through pattern matching, PPExtractor could effectively extract the privacy information collected and used by developers and third parties in privacy regulations, as well as the privacy information that Apps required users to input, providing a basis for management to regulate Android Apps' Illegal acts related to privacy information. This method could more comprehensively extract the privacy information collected in the Chinese privacy regulations. The experimental results show that PPExtractor can effectively identify privacy-related sentences in privacy policies and extract information related to user privacy in the sentences. [ABSTRACT FROM AUTHOR]

The refining of information from the immense amount of unstructured data on the internet can be a critical issue in identifying public opinion. It is difficult to extract relevant concepts from huge amounts of data. Concept level semantic parsing is improved over word-based investigation as it conserves the semantical data relevant to many-word articulations. The semantic proposals offer a superior comprehension of textual data and serve to altogether precision the exactness of numerous mining operations in text assignments. The extraction of concepts from textual data is a significant step forward in content analysis at the concept stage. We present a CLSA-CapsNet method that extracts concepts from natural language text. Then the extracted concepts are applied in Capsule networks (CapsNet). Moreover, the integration of Concept Level Sentiment Analysis (CLSA) and Capsule Networks (CapsNet) has not yet been implemented on the hotel review dataset. This is the first attempt, which we researched and embraced by the capsule network, to develop classification models for hotel reviews. The developed results demonstrated excellent performance with a prediction accuracy of 86.6% for CLSA-CapsNet models, respectively. Various similarities have also been made across our techniques and they are implemented by some other deep learning algorithms, such as rnn-lstm. Overall, the outstanding success obtained by CLSA-CapsNet in this investigation highlights its ability in the hotel review dataset. We likewise show exploratory outcomes, in which the proposed system outpaced the state-of-the-art CLSA-CapsNet model. [ABSTRACT FROM AUTHOR]

Efficient feature extraction and its application in scene labelling is a rapidly growing field in the area of computer vision and image processing. Many applications such as semantic labelling, object tracking, and augmentative reality have required robust feature selection methods. This paper proposes a two-step robust LM-MFP method for pixel-wise semantic image parsing in a pixel-labelled environment. In step 1, we have generated the morphological features from an image using attribute filters. Firstly, the image features based on attributes like area, diagonal, and inertia have been extracted and arranged in the form of a min-tree and max-tree architecture. Then, we have decided on some iterative threshold for tree pruning and obtained several distinct and redundant features. Although these features are highly informative, many are redundant. Therefore in step 2, we have explored a feature pooling method to discard the redundant features using similarity measures. For the measurement of similarity between two features, we have taken three parameters, i.e. correlation (ρ ), structural similarity (SSIM), and mutual information (MI). Subsequently, we formulated a hypothesis based upon the predicate containing ρ , SSIM, and MI as a variable. Based on the hypothesis, the redundant features are extracted, and a weighted average has taken for those features. This process has resulted in an exclusive and non-redundant features space. The reduced feature space has been used for pixel-based semantic classification for labelled images. Finally, we have applied a multi-class SVM to train the available pixel-label set model. The complete pixel set of the image has been used in the prediction of labelled objects by using the trained SVM model. The proposed LM-MFP approach has achieved 90.96, 89.03, 90.57, 91.12, and 93.63% accuracy for highway, house, sheep, horse rider, and horse keeper images, respectively, obtained from sift-flow and pascal-voc datasets. Conclusively in the comparative analysis section, the LM-MFP method has been observed to be state-of-art since it has significantly outperformed the previous techniques. [ABSTRACT FROM AUTHOR]

Existing deep models for facade parsing often fail in classifying pixels in heavily occluded regions of facade images due to the difficulty in feature representation of these pixels. In this paper, we solve facade parsing with occlusions by progressive feature learning. To this end, we locate the regions contaminated by occlusions via Bayesian uncertainty evaluation on categorizing each pixel in these regions. Then, guided by the uncertainty, we propose an occlusion-immune facade parsing architecture in which we progressively re-express the features of pixels in each contaminated region from easy to hard. Specifically, the outside pixels, which have reliable context from visible areas, are re-expressed at early stages; the inner pixels are processed at late stages when their surroundings have been decontaminated at the earlier stages. In addition, at each stage, instead of using regular square convolution kernels, we design a context enhancement module (CEM) with directional strip kernels, which can aggregate structural context to re-express facade pixels. Extensive experiments on popular facade datasets demonstrate that the proposed method achieves state-of-the-art performance. [ABSTRACT FROM AUTHOR]