Search

Voice conversion is a research topic in the fields of speech and artificial intelligence. The goal of voice conversion is to change the timbre of speech while preserving the content of the source speech, making it sounds like spoken by the target speaker. It is essential to ensure both the quality and naturalness of the converted speech. Voice conversion based on nonparallel data gains much attention currently, where models are trained using non-parallel multilingual speaker datasets, enabling many-to-many and any-to-any voice conversions. This paper provides a comprehensive summary and analysis of recent developments in non-parallel voice conversion. Firstly, we outline the early voice conversion techniques based on parallel corpus and their limitations. Then, we introduce and compare various approaches to voice conversion based on nonparallel data, providing a thorough analysis. Finally, a summary and outlook on voice conversion technology is provided. [ABSTRACT FROM AUTHOR]

In response to the existing gun sound recognition and positioning tasks, which require separate identification and positioning, resulting in time-consuming computation, system redundancy, and complex development processes, this paper proposes to use a two-stage CRNN deep learning network model to complete the gun sound recognition and positioning tasks. Firstly, perform a logarithmic Mel transform on the collected gunshot signal and calculate the generalized phase transition cross correlation spectrum as input to the network model. Secondly, in the first stage, the gunshot signal is identified through the CRNN network. Finally, in the second stage, the introduction of a mask is used to determine whether the CRNN network weight sharing is implemented for localization. The method proposed in this article can effectively solve the problems of separate recognition and positioning tasks, system redundancy, and complex development processes in traditional methods, and has certain application value in achieving joint recognition and positioning. [ABSTRACT FROM AUTHOR]

The intermittent, random and uncontrollable power generation characteristics of renewable energy pose challenges for the full utilization of green energy. The high energy consumption feature of the virtual power plant data center makes it an efficient absorption and regulation strategy for the intermittent (non-dispatchable) power in renewable energy. This paper proposes a method to predict the "source-load" dual-state of the virtual power plant using a long short-term memory network that incorporates time-embedded encoding. The results indicate that using the model presented in this paper can achieve proactive alerts for "power shortages" at 15-minute intervals, creating ample buffer time windows for container suspension and backup. Combined with container technology, it realizes dynamic energy consumption management in data centers, thereby enhancing the robustness of the virtual power plant data center against power supply-demand imbalances. This technology is of great significance for stabilizing grid operations, accelerating the application of green clean energy, constructing a service pattern for the energy ecosystem and speeding up the digital transformation of the grid. [ABSTRACT FROM AUTHOR]

The management of safety hazards in water conservancy engineering construction is transitioning towards informatization and intelligence. In order to efficiently mine valuable potential information from a large amount of unstructured construction safety hazard data, an intelligent recommendation system for construction safety hazard rectification based on deep learning is proposed. This paper is based on the TF-IDF algorithm to extract feature words of hidden danger, construct a safety hazard association Sankey diagram and display the information flow characteristics among construction sections, hazard features and hazard types. Then, this paper mines association rules in historical data based on the FP-Growth algorithm. In addition, the process of case retrieval recommendation is optimized by combining the sequence similarity matching algorithm and the Doc2Vec model. This paper uses 80 953 construction safety hazard information as the data source, which is recorded in the water resources allocation project of Pearl River Delta from 2019 to 2023. Example verification shows that the proposed method can match accurate rectification measures for current construction safety hazards, effectively assisting construction safety managers to identify and address hidden danger. [ABSTRACT FROM AUTHOR]

This paper analyzes and combs the application and research status of artificial intelligence technology in the field. of military intelligence, in order to provide reference for subsequent military intelligence research. This paper summarizes the development and application of artificial intelligence in military intelligence work from the aspects of intelligence analysis and military command decision-making. Based on the intelligence workflow, the military intelligence service model under artificial intelligence technology is analyzed. This paper systematically combs the research and development status of typical projects of intelligent intelligence systems in the United States, and analyzes the key development trends and technical difficulties of artificial intelligence in the military intelligence field. [ABSTRACT FROM AUTHOR]

[Purpose/Significance] The development of the Internet has led to the rapid development of social networks, providing users with a convenient channel for the release, dissemination and acceptance of information. However, its low-threshold characteristics have also given rise to a group of the "Internet water army"・・online social spammers, who are paid to post online comments with particular content and spread false information on purpose. They have become a major problem for the Internet ecology. It is of great significance to detect the Internet water army, prevent their malicious attacks, and combat and eliminate their negative effects on the security of the online public opinion. [Method/Process] First, we analyzed the development process and characteristics of the online social spammers, summarized the algorithms used in previous studies and the characteristics mentioned, and sorted out three research starting points: text features, interaction features and graph structure features. Then, an online social spammer detection method based on deep learning was proposed. Combined with the three aspects of user basic information, historical remarks and interaction behavior, six types of features were extracted from the basic information, recent remarks, social intimacy, interaction behavior, microblog number and membership level. Through feature depth extraction and vector splicing and fusion, the user feature vectors were formed with the same length. Finally, a convolutional neural network was used as a classifier to build an automatic, high-precision and high-efficiency spammer detection model. Two Chinese online spammer datasets collected from the Sina Weibo platform were selected for the experiment. The features of the datasets were spliced and aligned to form the Weibo Spammer 2023 dataset as the model training dataset, which prevented the data features of a single dataset from being too discrete and reducing modle generalization. Considering the overfitting problem in the model training process, we solved the problem by adding abandoned layers. [Results/Conclusions] The online spammer detection model constructed in this experiment has significantly improved in terms of metrics such as precision and accuracy. At the same time, the ablation experiment shows that the six features extracted in this e迎eriment have a positive effect on the detection process. Through empirical analysis, the model constructed in this paper has a high detection accuracy and detection efficiency, which can provide certain technical support and theoretical guidance for online spammer identification. By using machine learning methods to actively identify online social spammer accounts, real-time monitoring and prevention of key spammer accounts can prevent the occurrence of malicious network events more timely and effectively and reduce the risk of illegal forces damaging the public opinion ecology. [ABSTRACT FROM AUTHOR]

Current point cloud semantic segmentation methods based on deep learning tend to overlook the boundary of objects in transition area, resulting in the problem of ambiguous features at the boundary. This article proposed a point cloud semantic segmentation method with boundary-aware and multi-feature fusion (BA-MFF). Firstly, the backbone network was optimized to make the extracted features more robust. Secondly, boundary-aware module(BAM)was designed to focus on the boundaries of objects in transition area. This module consisted of a boundary point prediction module(BPPM)and a feature aggregation module (FAM). The boundary point prediction module predicted the points belonging to boundaries by learning the features of neighboring points, and the feature aggregation module performed discriminative aggregation of point cloud features within the neighborhood. Finally, for more discriminative features, a multi-feature fusion module(MFFM)was introduced, which fused features between different channels. The experimental results show that the mean Intersection over Union (mIoU) of this method reaches 63.7% on the ScanNetV2 dataset. The Overall Accuracy (OA) and mIoU on the S3DIS are 88.2% and 62.3%, respectively. The method in this paper effectively focuses on the transition area and has some segmentation superiority. [ABSTRACT FROM AUTHOR]

The traditional nonnegative matrix factorization methods are based on single-layer model. The deep nonnegative matrix factorization methods are based on the mathematical theory, ignoring the actual spectral mixing process of materials. This paper started from the physical process of spectral mixing, combining the advantages of traditional nonnegative matrix factorization models and deep learning. Aiming at the problem that the deep nonnegative matrix factorization methods ignore the actual spectral mixing process of materials, the spectral mixing process is reversed from the physical spectral mixing process. Furthermore, considering the sparsity and spatial smoothness of the distribution of the endmembers in the two-dimensional space, the total variation and sparsity regularized deep nonnegative matrix factorization based on the endmember matrix is proposed. This paper verified the effectiveness of the proposed method through synthetic experiments and real data experiments, which is compared with five classical hyperspectral unmixing methods. Compared with SDNMF-TV, SAD and RMSE are reduced. The results both in the synthetic experiments and the real experiments shows that the method in this paper achieves the best unmixing results. [ABSTRACT FROM AUTHOR]

[Objective] The root system is an important component of plant composition, and its growth and development are crucial for plants. Root image segmentation is an important method for obtaining root phenotype information and analyzing root growth patterns. Research on root image segmentation still faces difficulties, because of the noise and image quality limitations, the intricate and diverse soil environment, and the ineffectiveness of conventional techniques. This paper proposed a multi-scale feature extraction root segmentation algorithm that combined data augmentation and transfer learning to enhance the generalization and universality of the root image segmentation models in order to increase the speed, accuracy, and resilience of root image segmentation. [Methods] Firstly, the experimental datasets were divided into a single dataset and a mixed dataset. The single dataset acquisition was obtained from the experimental station of Hebei Agricultural University in Baoding city. Additionally, a self-made RhizoPot device was used to collect images with a resolution pixels of 10,200×14,039, resulting in a total of 600 images. In this experiment, 100 sheets were randomly selected to be manually labeled using Adobe Photoshop CC2020 and segmented into resolution pixels of 768× 768, and divided into training, validation, and test sets according to 7:2:1. To increase the number of experimental samples, an open source multi-crop mixed dataset was obtained in the network as a supplement, and it was reclassified into training, validation, and testing sets. The model was trained using the data augmentation strategy, which involved performing data augmentation operations at a set probability of 0.3 during the image reading phase, and each method did not affect the other. When the probability was less than 0.3, changes would be made to the image. Specific data augmentation methods included changing image attributes, randomly cropping, rotating, and flipping those images. The UNet structure was improved by designing eight different multi-scale image feature extraction modules. The module structure mainly included two aspects: Image convolution and feature fusion. The convolution improvement included convolutional block attention module (CBAM), depth wise separable convolution (DP Conv), and convolution (Conv). In terms of feature fusion methods, improvements could be divided into concatenation and addition. Subsequently, ablation tests were conducted based on a single dataset, data augmentation, and random loading of model weights, and the optimal multi-scale feature extraction module was selected and compared with the original UNet. Similarly, a single dataset, data augmentation, and random loading of mod‐ el weights were used to compare and validate the advantages of the improved model with the PSPNet, SegNet, and DeeplabV3Plus algorithms. The improved model used pre-trained weights from a single dataset to load and train the model based on mixed datasets and data augmentation, further improving the model's generalization ability and root segmentation ability. [Results and Discussions] The results of the ablation tests indicated that Conv_ 2+Add was the best improved algorithm. Compared to the original UNet, the mIoU, mRecall, and root F1 values of the model increased by 0.37%, 0.99%, and 0.56%, respectively. And, comparative experiments indicate Unet+Conv_2+Add model was superior to the PSPNet, SegNet, and DeeplabV3Plus models, with the best evaluation results. And the values of mIoU, mRecall, and the harmonic average of root F1 were 81.62%, 86.90%, and 77.97%, respectively. The actual segmented images obtained by the improved model were more finely processed at the root boundary compared to other models. However, for roots with deep color and low contrast with soil particles, the improved model could only achieve root recognition and the recognition was sparse, sacrificing a certain amount of information extraction ability. This study used the root phenotype evaluation software Rhizovision to analyze the root images of the Unet+Conv_2+Add improved model, PSPNet, SegNet, and DeeplabV3Plu, respectively, to obtain the values of the four root phenotypes (total root length, average diameter, surface area, and capacity), and the results showed that the average diameter and surface area indicator values of the improved model, Unet+Conv_2+ Add had the smallest differences from the manually labeled indicator values and the SegNet indicator values for the two indicators. Total root length and volume were the closest to those of the manual labeling. The results of transfer learning experiments proved that compared with ordinary training, the transfer training of the improved model UNet+Conv_2+Add increased the IoU value of the root system by 1.25%. The Recall value of the root system was increased by 1.79%, and the harmonic average value of F1 was increased by 0.92%. Moreover, the overall convergence speed of the model was fast. Compared with regular training, the transfer training of the original UNet improved the root IoU by 0.29%, the root Recall by 0.83%, and the root F1 value by 0.21%, which indirectly confirmed the effectiveness of transfer learning. [Conclusions] The multi-scale feature extraction strategy proposed in this study can accurately and efficiently segment roots, and further improve the model's generalization ability using transfer learning methods, providing an important research foundation for crop root phenotype research. [ABSTRACT FROM AUTHOR]

[Purpose/Significance] Deep learning for language modeling is one of the major methods and advanced technologies to enhance language intelligence of machines at present, which has become an indispensable important technical means for automatic processing and analysis of data resources, and intelligent mining of information and knowledge. However, there are still some difficulties in using deep learning for language modeling for technology development and application service in the library and information science (LIS) field. Therefore, this study systematically reviews and reveals the research progress, technical principles, and development methods of deep learning for language modeling, with the aim at providing reliable theoretical basis and feasible methodological paths for the deep understanding and application of deep learning for language modeling for librarians and fellow practitioners. [Method/Process] The data used in this study were collected from the WOS core database, CNKI literature database, arXiv preprint repository, GitHub open-source software hosting platform and the open resources on the Internet. Based on these data, this paper first systematically investigates the background, basic feature representation algorithms, and representative application development tools of deep learning for language modeling, reveals their dynamic evolution and technical principles, and analyzes the advantages and disadvantages and applicability of each algorithm model and development tool. Second, an in-depth analysis of the possible challenging problems faced by the development and application of deep learning for language modeling was performed, and two strategic approaches to expand their application capabilities were put forward. [Results/Conclusions] The important challenges faced by the application and development of deep learning for language modeling include numerous parameters and difficulties to adjust accuracy, relying on a large amount of accurate training data, difficulties in making changes, and the intellectual property and information security issues. In the future, we will start from two aspects of specific domains and feature engineering to expand and improve the application capabilities of deep learning for language modeling. Specifically, we focus on consideration of the collection and preparation of domain data, selection of model architecture, participation of domain experts, and optimization for specific tasks, in order to ensure that the data source of the model is more reliable and secure, and the application effect is more accurate and practical. Moreover, the strategic methods for feature engineering to expand the application capabilities of deep learning for language modeling include selecting appropriate features, feature pre-processing, feature selection, and feature dimensionality reduction. These strategies can help improve the performance and efficiency of deep learning for language models, making them more suitable for specific tasks or domains. To sum up, LIS institutions should leverage the deep learning for language modeling related technologies, guided by the needs of scientific research and social development, and based on advantages of existing literature data resources and knowledge services; they should carry out innovative professional or vertical domain intelligent knowledge management and application service, and develop technology and systems with independent intellectual property rights, which is their long-term sustainable development path. [ABSTRACT FROM AUTHOR]

[Purpose/Significance] Advances in single-cell sequencing and high-throughput technology have made it possible for plant genomics to accumulate large quantities of data describing multidimensional genomic-wide molecular phenotypes at low cost. As powerful data mining tools, deep learning techniques can be utilized to further predict and interpret the acquired molecular phenotypes. In recent studies, deep learning has been shown to yield significant results in plant genomics and crop breeding research. However, a complete review of deep learning applications in plant genomics is lacking. [Method/Process] The input to deep learning applied to genomics is usually biological sequences and molecular phenotypes as predictor and target variables, respectively. We introduced the workflow from four views: input data pre-processing includes retrieval, coding, and splitting; model construction and training includes the selection of model architecture and hyperparameters; model evaluation and interpretability. Specifically, this paper introduces the background of deep learning approaches, including the latest graph neural networks; then it discusses two prominent issues in the intersection of genomics and deep learning with respect to gene characterization and protein characterization: 1) how to model the flow of information from plant genomic DNA sequences to molecular phenotypes; and 2) how deep learning models can be utilized to identify functional variation in natural populations? Specifically, the paper summarizes the current status of deep learning applications in related fields, which include deep learning and DNA and gene characterization research, interpretability of deep learning in genomics applications, graph neural networks in genomics, deep learning and genomic variation research, deep learning in protein prediction, ALPHAFOLD in protein prediction, deep learning and crop breeding research, and unsupervised learning in genomics and protein characterization. [Results/Conclusions] This article summarizes how traditional deep-learning algorithms, graph deep-learning, generative adversarial networks and interpretable AI are applied in current research in order to address these two problems. Finally, the prospects for deep learning in future plant genomics research and crop improvement are discussed. Overall, deep learning has provided better results than conventional methods in many genomics research directions, and the application of deep learning in genomics has yielded early applications of scientific and economic significance. Deep learning offers two distinct advantages: 1) end-to-end learning, with the ability to integrate multiple pre-processing steps into a single model; and 2) multimodal data processing capabilities that can handle extremely heterogeneous data in genomics. The advancement of deep learning has the potential to expand new research perspectives in genomics and crop breeding, and to facilitate larger-scale association studies in both phenotypic and genotypic genomics as algorithms become more accurate. [ABSTRACT FROM AUTHOR]

Street vendors are an indispensable part of the urban social ecosystem, but due to a lack of comprehensive understanding, many cities have adopted simple eviction policies, resulting in the gradual marginalization and stigmatization of the street economy. The efficient governance of street vendors requires the comprehensive investigation of their business scale and spatial distribution information. However, traditional methods have limitations in terms of automatically surveying large-scale street vendor information, particularly spatial distribution. This paper proposes a method for the automatic investigation spatial distribution of street vendors based on street-view images and a deep-learning object recognition model. Street-view images were collected at fixed intervals according to the urban road network, and 1, 957 images containing one or more vendors were selected through human-machine interaction to establish street vendor label data. To achieve high recognition model accuracy, the category labels were subdivided into four categories: ground stalls, table stalls, tricycle stalls, and small truck stalls, based on the goods carriers used by street vendors. A deep neural-networkbased image object detection model based on YOLO v4 was constructed to identify street vendors in the streetview image library, with an average F1 value of 0.77 and an mAP of 0.67. The accuracy of the model was satisfactory for investigating the number and location of street vendors covering the main roads in the city and then applying a kernel density distribution model to evaluate the spatial distribution pattern of street vendors. Using street vendors in Guangzhou as a case study, the proposed automatic investigation model identified 26,119 street vendors from 3,339,062 street-view images. The results showed that the street vendors were distributed in a multicenter aggregation pattern in the central urban area, mainly concentrated in areas with high pedestrian traffic, such as subway stations and urban villages; their numbers increased as road grades decreases. Street vendors were mainly distributed in areas with medium rents. The proposed method is helpful for performing the efficient, low-cost, and city-scale mapping of street vendors; the results obtained provide suggestions for formulating and implementing spatial governance policies for the informal economy and further provide suggestions for improving and implementing spatial governance policies for open and diverse urban street-view images. The results can be used as a reference for the location preference analysis of practitioners, the exploration of NIMBY syndrome, and the determination of the formalization zone. Although street-view images have an insufficient spatiotemporal coverage, using them to perform street vendor investigations is a low-cost and efficient method compared with the use of traditional investigation methods and data sources. In addition, the method proposed in this article can be coupled with multitask deep learning algorithms to investigate additional dimensions of street vendor information, such as the sex, age, and type of business of street vendors. Relevant research needs to be conducted in the future. [ABSTRACT FROM AUTHOR]

[Purpose/Significance] Abstracts of scientific documents are often composed of sections with specific functions. Using the deep learning method to identify structural functions of abstracts of scientific documents is conducive to the in-depth analysis of the documents. [Method/Process] In this paper, identifying structural functions of abstracts of scientific documents is transformed into a text classification problem, and its structure functions are divided into four categories: "introduction, methods, results, conclusions (IMRC)". Based on the text content and context features of abstract sentences, the classifier is constructed based on deep learning models such as BERT, BERT-BiLSTM, BERT-TextCNN and ERNIE, to automatically identify structural functions of abstracts of scientific documents. [Results/Conclusions] Experiments are carried out on a dataset with 3,130 articles in the field of eHealth. The results show that the scores of indicators for ERNIE are higher than other models. BERT-TextCNN model is better in dealing with short text, while BERT-BiLSTM model is better in handling long sentences. The method proposed in this paper is helpful for the fine-grained functional understanding of scientific literature abstracts, and is of great significance to the in-depth mining of scientific literature and literature based knowledge discovery. [ABSTRACT FROM AUTHOR]

Deep learning has gone through breakthroughs in many research fields including computer vision, natural language processing, etc. due to multiple driving factors such as knowledge, data, algorithms and computing power. In addition, it has gradually spawned a number of new research directions with the migration and application as well as cross-integration among various disciplines. Cheminformatics is a discipline that solves chemical problems with the applied informatics methods, and deep learning can be useful since it is very powerful in nonlinear learning. Deep learning model can be used to screen and predict in the data set, and then verify the feasibility of the results based on theoretical calculation. Finally, the results are represented by experiments, which shortens the experimental period, reduces the labor cost and accelerates the intelligence of cheminformatics. This paper briefly introduces the development history and main network model architecture of deep learning as well as the latest research and application status of deep learning in synthesis planning, compound structure-activity relation and catalyst design in recent years, and also discusses and expects the future development direction. [ABSTRACT FROM AUTHOR]

Accurate and efficient monitoring of animal information, timely analysis of animal physiological and physical health conditions, and automatic feeding and farming management combined with intelligent technologies are of great significance for large-scale livestock farming. Deep learning techniques, with automatic feature extraction and powerful image representation capabilities, solve many visual challenges, and are more suitable for application in monitoring animal information in complex livestock farming environments. In order to further analyze the research and application of artificial intelligence technology in intelligent animal farming, this paper presents the current state of research on deep learning techniques for tag detection recognition, body condition evaluation and weight estimation, and behavior recognition and quantitative analysis for cattle, sheep and pigs. Among them, target detection and recognition is conducive to the construction of electronic archives of individual animals, on which basis the body condition and weight information, behavior information and health status of animals can be related, which is also the trend of intelligent animal farming. At present, intelligent animal farming still faces many problems and challenges, such as the existence of multiple perspectives, multi-scale, multiple scenarios and even small sample size of a certain behavior in data samples, which greatly increases the detection difficulty and the generalization of intelligent technology application. In addition, animal breeding and animal habits are a long-term process. How to accurately monitor the animal health information in real time and effectively feed it back to the producer is also a technical difficulty. According to the actual feeding and management needs of animal farming, the development of intelligent animal farming is prospected and put forward. First, enrich the samples and build a multi perspective dataset, and combine semi supervised or small sample learning methods to improve the generalization ability of in-depth learning models, so as to realize the perception and analysis of the animal's physical environment. Secondly, the unified cooperation and harmonious development of human, intelligent equipment and breeding animals will improve the breeding efficiency and management level as a whole. Third, the deep integration of big data, deep learning technology and animal farming will greatly promote the development of intelligent animal farming. Last, research on the interpretability and security of artificial intelligence technology represented by deep learning model in the breeding field. And other development suggestions to further promote intelligent animal farming. Aiming at the progress of research application of deep learning in livestock smart farming, it provides reference for the modernization and intelligent development of livestock farming. [ABSTRACT FROM AUTHOR]

Ecological conservation red line is an important initiative for ecological civilization construction in China. Studies and works have been conducted to delineate the ecological conservation red line using remote sensing and manual survey methods. However, the traditional ecological assessment methods ignore the impact of dynamic elements on the ecological environment, and a permanent monitoring system for ecological conservation red line is still necessary. Therefore, this paper aimed to build a daily monitoring system of the ecological conservation red line from the perspective of ecological soundscape, combining deep learning and traditional Geographic Information System (GIS) methods. We analyzed the main factors affecting the ecological soundscape to provide a benchmark for selecting sites for permanent soundscape monitoring. Initially, this study developed the first classification standard of soundscape elements for daily monitoring of ecological conservation red line concerning the "ecological-production-living spaces" concept and existing literature on soundscape and built a deep learning model for training, with training and validation accuracies of 89.88% and 72.41%, respectively. The identification results of this model confirmed that the solution of intelligent monitoring through soundscape is reliable in daily monitoring. Then, the deep learning model was employed to classify and predict the soundscape collected from the case sites and effectively identify the areas with severe ecological encroachment by calculating the proportion of ecological soundscape elements, performing the simulation of daily monitoring, which is important guidance for the actual ecological conservation red line monitoring. Finally, a regression model was used to analyze the spatial characteristics of ecologically vulnerable areas, with an R2 of 0.472. The regression model showed that nighttime light intensity had the greatest influence on ecological soundscape, followed by the distance from residential areas, spatial centrality, and road distance. These factors should be considered when implementing permanent soundscape monitoring sites. In addition, Normalized Difference Vegetation Index (NDVI) and topography were not significant in the regression model, which confirmed the shortcomings of considering only the ecological background and ignoring the dynamic biological elements. Daily monitoring should capture dynamic encroachments so that major ecological damage can be detected before it occurs. Therefore, the innovative integration of soundscape elements into the monitoring system meets the basic needs of monitoring ecological conservation red lines daily. [ABSTRACT FROM AUTHOR]

[Purpose/Significance] At present, most of the training models used in the research of news classification are non-active learning. There are common problems about these models, including data cannot be labeled immediately and the labeling cost is too high, which also hinders the analysis of agricultural news. Especially because of the explosive growth of news data in the network era, it is more difficult to label data, train supervised text classification models, and screen relevant news in the field of agriculture from diversified online news sources. In order to solve this problem, the most commonly used pool based active learning or deep active learning technique is used to select more valuable and representative data from unlabeled data for manual labeling, and construct labeled data sets to improve the efficiency and effect of news classification and agricultural news mining. [Method/Process] The commonly used machine learning models for text classification, such as random forest classifier, polynomial naive Bayes classifier and logistic regression classifier, were combined with the active learning method with the lowest confidence to analyze the effect, and the BERT model was combined with the three sampling strategies of discriminative active learning, deep Bayes active learning and lowest confidence for deep active learning training. On the news corpus of 19 847 samples crawled and cleaned by crawler technology from Sina and other news websites, aiming at screening agricultural related news from diversified news samples of various topics, the iterative experiment of adding 30 samples per round was tested to check the improvement effect of F1 score under various method combinations with the increase of the number of annotation. In addition, the representativeness and diversity of the samples selected by the sampling function of each method in the deep active learning method of the BERT model were compared, so as to understand the characteristics of each strategy and provide inspiration for the selection and improvement of Al strategy in the future. In addition, this paper also analyzed how much labeling cost can be saved by using the proposed method. [Results/Conclusions] When comparing a variety of machine learning models, it is found that although the gradient boosting tree and support vector machine classifier have high accuracy, they are not suitable for active learning because of their low efficiency in text data processing of large-scale high-dimensional data. After combining other machine learning models and the BERT model and training text models with the corresponding active learning or deep active learning methods, it is found that the application of active learning method can significantly improve the training process of each model. Among them, the BERT model, combined with discriminative active learning sampling function, has the best news text classification effect and the lowest annotation data requirements. The representativeness and diversity of the samples selected by discriminative active learning sampling function are also the highest, which explains the source of the advantages of this method. It can also be found that for the same task model, the higher the accuracy of classification is required, and the active learning method can save more annotation cost than non-active learning. [ABSTRACT FROM AUTHOR]

Non-small cell lung cancer (NSCLC) accounts for 80%-85% of lung cancer and is one of the malignant tumors that seriously endanger human health. Treatment of early NSCLC is based on surgical excision or stereotactic body radiation therapy. Whether there is regional lymphatic metastasis when diagnosis is confirmed will affect the choice of local treatment, and whether there is still a risk of lymphatic and hematologic metastasis after the completion of local therapy will be the basis for accurate decision of adjuvant therapy. How to predict the risk of lymphatic or hematologic metastasis of NSCLC remains a challenge. With the development of tumor and the plasticity of treatment, heterogeneity of biological characteristics of tumors in time and space seriously affects the accuracy of clinical diagnosis, treatment and prognostic prediction. Due to the heterogeneity of tumor, it is difficult for invasive biopsy to show the full picture of tumor biological characteristics as a gold standard, which promotes clinical attention to non-invasive methods, such as medical images, to identify biological features. The method to identify tumor biological features based on medical images experiences from the qualitative analysis of artificial visuals to the modeling of advanced statistical methods for manual extraction of imaging features, and then to the application of radiomics and deep learning models, which provide new possibilities for accurate and efficient medical imaging analysis. Based on chest computed tomography (CT) imaging, this paper summarized the progress of research on the prediction of risk of lymphatic and hematologic metastasis, an important factor affecting early NSCLC treatment decision-making. [ABSTRACT FROM AUTHOR]

Accurate detection and recognition of plant diseases is the key technology to early diagnosis and intelligent monitoring of plant diseases, and is the core of accurate control and information management of plant diseases and insect pests. Deep learning can overcome the disadvantages of traditional diagnosis methods and greatly improve the accuracy of diseases detection and recognition, and has attracted a lot of attention of researchers. This paper collected the main public plant diseases image data sets all over the world, and briefly introduced the basic information of each data set and their websites, which is convenient to download and use. And then, the application of deep learning in plant disease detection and recognition in recent years was systematically reviewed. Plant disease target detection is the premise of accurate classification and recognition of plant disease and evaluation of disease hazard level. It is also the key to accurately locate plant disease area and guide spray device of plant protection equipment to spray drug on target. Plant disease recognition refers to the processing, analysis and understanding of disease images to identify different kinds of disease objects, which is the main basis for the timely and effective prevention and control of plant diseases. The research progress in early disease detection and recognition algorithm was expounded based on depth of learning research, as well as the advantages and existing problems of various algorithms were described. It can be seen from this review that the detection and recognition algorithm based on deep learning is superior to the traditional detection and recognition algorithm in all aspects. Based on the investigation of research results, it was pointed out that the illumination, sheltering, complex background, different disorders with similar symptoms, different changes of disease symptoms in different periods, and overlapping coexistence of multiple diseases were the main challenges for the detection and recognition of plant diseases. At the same time, the establishment of a large-scale and more complex data set that meets the specific research needs is also a difficulty that need to face together. And at further, we point out that the combination of the better performance of the neural network, large-scale data set and agriculture theoretical basis is a major trend of the development of the future. It is also pointed out that multimodal data can be used to identify early plant diseases, which is also one of the future development direction. [ABSTRACT FROM AUTHOR]

Diseases and pests are main stresses to crop production. It is necessary to accurately and quickly monitor and control the stresses dynamically, so as to ensure the food security and the quality and safety of agricultural products, protect the ecological environment, and promote the sustainable development of agriculture. In recent years, with the rapid development of the unmanned aerial vehicle (UAV) industry, UAV agricultural remote sensing has played an important role in the application of crop diseases and pests monitoring due to its high image spatial resolution, strong data acquisition timeliness and low cost. The relevant background of UAV remote sensing monitoring of crop disease and pest stress was introduced, then the current methods commonly used in remote sensing monitoring of crop disease and pest stress by UAV was summarized. The data acquisition method and data processing method of UAV remote sensing monitoring of crop disease and pest stress were mainly discussed. Then, from the six aspects of visible light imaging remote sensing, multispectral imaging remote sensing, hyperspectral imaging remote sensing, thermal infrared imaging remote sensing, LiDAR imaging remote sensing and multiple remote sensing fusion and comparison, the research progress of remote sensing monitoring of crop diseases and pests by UAV worldwide was reviewed. Finally, the unresolved key technical problems and future development directions in the research and application of UAV remote sensing monitoring of crop disease and pest stress were proposed. Such as, the performance of the UAV flight platform needs to be optimized and upgraded, as well as the development of low-cost, lightweight, modular, and more adaptable airborne sensors. Convenient and automated remote sensing monitoring tasks need to be designed and implemented, and more remote sensing monitoring information can be obtained. Data processing algorithms or software should be designed and developed with greater applicability and wider applicability, and data processing time should be shortened by using 5G-based communication networks and edge computing devices. The applicability of the algorithm or model for UAV remote sensing monitoring of crop disease and pest stress needs to be stronger, so as to build a corresponding method library. We hope that this paper can help Chinese UAV remote sensing monitoring of crop diseases and pests to achieve more standardization, informatization, precision and intelligence. [ABSTRACT FROM AUTHOR]

Due to the various species of fish, and the influence of different light and background environments, the recognition accuracy of some traditional fish recognition algorithms bases on color texture or feature point extrac tion is reduced, and good classification results cannot be achieved. To solve the problem, based on the existing Alex Net convolutional neural network, this paper proposes one method to speed up model training by reducing redundant convolutional layers. An item-based soft attention algorithm is applied to the improved AlexNet convolutional neural network model, which consists of four convolutional layers, one item-based soft attention layer, and two fully-con nected layers. Meanwhile, a fish recognition model is established using transfer learning methods. The test results show that the average recognition accuracy of the proposed algorithm achieves 97. 43%, which is 4.08% higher than the original AlexNet model, and the average recognition rate of some fishes achieves 99.31% and time consumption of fish recognition is reduced by 35%. In conclusion, compared with state-of-the-art fish recognition algorithms. the fish recognition algorithm proposed in this study achieve higher accuracy, lower model complexity, and stronger robustness. [ABSTRACT FROM AUTHOR]

Medical image segmentation plays an important role in clinical diagnosis and is the basis of other medical image processing methods. With the improvement of computer hardware performance, image segmentation technology based on deep learning has already become a powerful tool for processing medical images and is widely used in various medical image segmentation tasks. This paper introduces several types of common medical images and their characteristics, analyzes and compares the image segmentation algorithms that have emerged in recent years. Some algorithms have been successfully applied to segmentation tasks such as brain tissue, lungs and blood vessels. In response to the current problems in the development of medical image segmentation technology based on deep learning, corresponding strategies are proposed, and the future development direction is also prospected. [ABSTRACT FROM AUTHOR]

Musculoskeletal diseases are diverse. Medical imaging examination is the primary means of disease diagnosis. Reasonable selection of image post-processing methods can provide reliable evaluation basis for disease diagnosis, surgery and prognosis. As a new stage of the development of computer science, artificial intelligence technology can efficiently and accurately preprocess and analyze images, and assist clinicians in diagnosis and treatment. It will bring new opportunities for the development of musculoskeletal imaging. This paper reviews the current situation and challenges of artificial intelligence in the field of imaging diagnosis of musculoskeletal system diseases, in order to provide some reference for the relevant research scholars. [ABSTRACT FROM AUTHOR]