Search

In the realm of biological image analysis, deep learning (DL) has become a core toolkit, for example for segmentation and classification. However, conventional DL methods are challenged by large biodiversity datasets characterized by unbalanced classes and hard‐to‐distinguish phenotypic differences between them. Here we present BioEncoder, a user‐friendly toolkit for metric learning, which overcomes these challenges by focussing on learning relationships between individual data points rather than on the separability of classes. BioEncoder is released as a Python package, created for ease of use and flexibility across diverse datasets. It features taxon‐agnostic data loaders, custom augmentation options, and simple hyperparameter adjustments through text‐based configuration files. The toolkit's significance lies in its potential to unlock new research avenues in biological image analysis while democratizing access to advanced deep metric learning techniques. BioEncoder focuses on the urgent need for toolkits bridging the gap between complex DL pipelines and practical applications in biological research. [ABSTRACT FROM AUTHOR]

Gaining large-scale, temporally continuous surface soil moisture information through remote sensing techniques is of significant importance for crop irrigation management and farmland drought early warning. The commonly used remote sensing soil moisture inversion method often utilizes the modified perpendicular dryness index (MPDI) in conjunction with ground-measured data to construct models. However, this method lacks universality, as each temporal phase of the image inversion requires the acquisition of corresponding ground-measured data for calibration, resulting in a high dependence on ground-measured data. To reduce the dependence on ground-measured data when estimating soil moisture using optical remote sensing data, this study firstly analyzed the feature space differences and the variation laws of soil line slopes using 54 temporal phases of Sentinel-2 T50SKH scenes from 2020 to 2021. The study also derived and explained the influence of soil line slope variations on soil moisture inversion, revealing the fundamental reason for the need for ground-measured data calibration for each period in the MPDI soil moisture inversion. Based on this, a multi-temporal soil line consistency correction method and a re-modified perpendicular drought index (RPDI) were proposed. Soil moisture inversion was conducted using RPDI, and the effectiveness was validated in the study area. The main conclusions are as follows: 1) The analysis of the feature space of the Sentinel-2 satellite's continuous two-year scenes indicated that the feature spaces of most temporal phases were incomplete. Theoretically existing completely bare or extremely wet points may not be present in a single image, leading to long-term fluctuations in soil line slopes across different temporal phases. For the modified perpendicular dryness index (MPDI), the soil lines used for the calculation of MPDI in different temporal phases were different, which could only represent the relative dryness or wetness of each phase. This was the fundamental reason why ground-measured data calibration was required for each temporal phase when using MPDI for soil moisture inversion. 2) Through theoretical derivation of the MPDI changes caused by soil line slope variations, it was found that the changes in the inversion equation's slope were mainly to offset the fluctuations in soil line slopes. For the same region, since soil texture did not change over time, if a stable soil line slope could be obtained, then the slope of the linear regression equation between soil moisture and MPDI would also remain stable, and calibration would not be necessary for each temporal phase. 3) A soil line consistency correction method and RPDI were proposed. Since the calculation of this index was based on a unified theoretical soil line, it stabilized the regression relationship between RPDI and soil moisture. Validation results showed that a unified regression equation could be used between RPDI and ground-measured soil moisture (the unbiased root mean square error of the calibration set was 2.88%, R² was 0.49, and the unbiased root mean square error of the validation set was 3.05%, R² was 0.54). When applying the regression equation obtained from the calibration set to other temporal phases, it was not significantly different from the MPDI regression equations established for each phase separately in terms of R² and unbiased root mean square error. This indicated that RPDI no longer depended on ground-measured data and could achieve direct comparison across different temporal phases, significantly reducing the dependence on ground-measured data for multi-temporal soil moisture estimation using optical remote sensing. In conclusion, this research offered a substantial contribution to the utilization of optical remote sensing data for large-scale, continuous soil moisture inversion. The introduced RPDI method not only simplified the process of soil moisture estimation but also fortified the resilience of soil moisture retrieval. Furthermore, it significantly boosted the efficiency and precision of soil moisture monitoring. As a result, it held great value for the implementation of sound irrigation strategies and the timely issuance of drought warnings. [ABSTRACT FROM AUTHOR]

In this article, we formalize the problem of semiautomatic construction of the contour of area objects from satellite hyperspectral images and present a solution algorithm using PCA and Dijkstra's algorithm. The contour is considered as the boundary of an object, which can be used for its segmentation and classification. The semiautomatic contour accepts reference points specified by the operator. The formalization of the algorithm is completed. [ABSTRACT FROM AUTHOR]

As a new vegetation monitoring index, the KNDVI has certain advantages in characterizing the evolutionary process of regional desertification. However, there are few reports on desertification monitoring based on KNDVI and feature space models. In this study, seven feature parameters, including the kernel normalized difference vegetation index (KNDVI) and Albedo, were introduced to construct different models for desertification remote-sensing monitoring. The optimal desertification remote-sensing monitoring index model was determined with the measured data; then, the spatiotemporal evolution pattern of desertification in Gulang County from 2013 to 2023 was analyzed and revealed. The main conclusions were as follows: (1) Compared with the NDVI and MSAVI, the KNDVI showed more advantages in the characterization of the desertification evolution process. (2) The point–line pattern KNDVI-Albedo remote-sensing index model had the highest monitoring accuracy, reaching 94.93%, while the point–line pattern NDVI-TGSI remote-sensing monitoring index had the lowest accuracy of 54.38%. (3) From 2013 to 2023, the overall desertification situation in Gulang County showed a trend of improvement with a pattern of "firstly aggravation and then alleviation." Additionally, the gravity center of desertification in Gulang County first shifted to the southeast and then to the northeast, indicating that the northeast's aggravating rate of desertification was higher than in the southwest during the period. (4) From 2013 to 2023, the area of stable desertification in Gulang County was the largest, followed by the slightly weakened zone, and the most significant transition area was that of extreme desertification to severe desertification. The research results provide important decision support for the precise monitoring and governance of regional desertification. [ABSTRACT FROM AUTHOR]

Exploring the complex meanings in architectural scenic images is crucial for modern intelligent systems. We propose a novel method combining multi-channel perceptual features to semantically understand scenes with intricate spatial structures. Our deep hierarchical model captures human gaze dynamics using a binary objectness measure to detect objects and details at various architectural scales. We introduce a multi-layer active learning paradigm to derive gaze shifting paths (GSPs) and calculate corresponding deep representations, tolerant to label noise via a penalty term that discards irrelevant GSP features. A manifold-regularized feature selector is used to choose high-quality GSP features, and a linear classifier distinguishes architectural scenes. Our technique outperforms traditional models, as shown in tests on three image retrieval datasets, demonstrating the discriminative capacity of our deep gaze-guided features.

The forest is an important part of carbon resources. Forest growing stock volume (GSV) is an important parameter of forest. The Water Cloud Model (WCM) is a simple equation that describes the interaction between ground objects and electromagnetic waves. It has also been applied in the estimation of forest GSV. When estimating GSV, the WCM equation parameters are usually calculated using least squares, but the least squares method relies on field reference data. The subsequent WCM development algorithm BIOMASAR uses a sliding window method that does not rely on measured data. However, the sliding window method is inefficient and can easily lead to missing pixels. We designed the backscatter/fractional vegetation coverage (FVC) feature space based on WCM and BIOMASAR to estimate forest GSV. Comparing with the national forest inventory (NFI) reference dataset and the BIOMASAR algorithm results in the study area, the method is evaluated from three aspects: accuracy, efficiency, and texture. The results show that this method does not rely on actual reference data, and the efficiency is increased from 1661s in the sliding window to 663s. The correlation with the NFI reference data is 0.45, the RMSE is 116.2327 m3/ha, and the RRMSE is 64.86%. The accuracy is better than the BIOMASAR sliding window GSV results in this study area, and compared with Google Earth images of the same period, it is also more consistent with the field texture. In short, the backscatter/FVC feature space can efficiently obtain forest GSV estimates more consistent with field conditions without relying on measured data.

As a new vegetation index, kernel normalized difference vegetation index (KNDVI) has great advantages in monitoring regional land degradation and vegetation status. However, the research on the spatial monitoring of rocky desertification based on KNDVI and feature space model has not been reported. In this study, the KNDVI, MSAVI, NDVI, EVI and BLI were introduced to establish four feature space monitoring index. After accuracy validation and comparisons, the optimal rocky desertification monitoring model was proposed and the spatio-temporal changes of rocky desertification had been analyzed. The time scale of this study was from 2018 to 2022, and the study area was about 3412 km2. The results showed that: (1) The rocky desertification monitoring model of KNDVI-BLI had the highest accuracy with R2=0.909 and RMSE of 0.392. (2) The rocky desertification in the western, northern and southeastern parts of Qixingguan District was more severe than other parts. (3) During 2018–2022, the rocky desertification showed an exacerbating trend in Qixingguan District. The area of extremely severe rocky desertification increased by 8.24 km2, while that of severe rocky desertification increased by 20.46 km2. The research results could provide a scientific and effective monitoring method for the control of rocky desertification.

Using the kernel normalized difference vegetation index (KNDVI) to monitor soil salinization has great advantages; however, approaches using KNDVI and a feature space model to monitor salinization have not yet been reported. In this study, the KNDVI, normalized difference vegetation index (NDVI), extended difference vegetation index (EDVI), green normalized difference vegetation index (TGDVI), modified soil-adjusted vegetation index (MSAVI), and salt index (SI) were used to establish five feature space monitoring indices for salinization. The spatio-temporal evolution pattern of soil salinization in the Yellow River Delta from 2000 to 2020 was analyzed based on the optimal monitoring index. The remote sensing monitoring index model based on KNDVI-SI's point-to-point mode had the best applicability with R2 = 0.93, followed by EDVI-SI's salinization monitoring index model with R2 = 0.90. From 2000 to 2020, soil salinization in the Yellow River Delta followed an exacerbating then improving trend. Soil salinization was more severe in the northern and eastern coastal areas of the Yellow River Delta. These results are conducive to salinization restoration and control in the Yellow River Delta. [ABSTRACT FROM AUTHOR]

Split learning is a widely recognized distributed learning framework suitable for joint training scenarios with limited computing resources. However, recent research indicates that the malicious server can achieve high-quality reconstruction of the client's data through feature space hijacking attacks, leading to severe privacy leakage concerns. In this paper, we further enhance this attack to enable efficient data reconstruction while maintaining acceptable performance on the main task. Another significant advantage of our attack framework lies in its ability to fool the state-of-the-art attack detection mechanism, thus minimizing the risk of attacker exposure and making sustainable attacks possible. Moreover, we adaptively refine and adjust the attack strategy, extending the data reconstruction attack for the first time to the more challenging scenario of vertically partitioned data in split learning. In addition, we introduce three training modes for the attack framework, allowing the attacker to choose according to their requirements freely. Finally, we conduct extensive experiments on three datasets and evaluate the attack performance of attack frameworks in different scenarios, parameter settings, and defense mechanisms. The results demonstrate our attack framework's effectiveness, invisibility, and generality. Our research comprehensively highlights the potential privacy risks associated with split learning and sounds the alarm for secure applications of split learning. [ABSTRACT FROM AUTHOR]

In recent years, various adversarial defense methods have been proposed to improve the robustness of deep neural networks. Adversarial training is one of the most potent methods to defend against adversarial attacks. However, the difference in the feature space between natural and adversarial examples hinders the accuracy and robustness of the model in adversarial training. This paper proposes a learnable distribution adversarial training method, aiming to construct the same distribution for training data utilizing the Gaussian mixture model. The distribution centroid is built to classify samples and constrain the distribution of the sample features. The natural and adversarial examples are pushed to the same distribution centroid to improve the accuracy and robustness of the model. The proposed method generates adversarial examples to close the distribution gap between the natural and adversarial examples through an attack algorithm explicitly designed for adversarial training. This algorithm gradually increases the accuracy and robustness of the model by scaling perturbation. Finally, the proposed method outputs the predicted labels and the distance between the sample and the distribution centroid. The distribution characteristics of the samples can be utilized to detect adversarial cases that can potentially evade the model defense. The effectiveness of the proposed method is demonstrated through comprehensive experiments. [ABSTRACT FROM AUTHOR]

To support pathologists in breast tumor diagnosis, deep learning plays a crucial role in the development of histological whole slide image (WSI) classification methods. However, automatic classification is challenging due to the high-resolution data and the scarcity of representative training data. To tackle these limitations, we propose a deep learning-based breast tumor gigapixel histological image multi-classifier integrated with a high-resolution data augmentation model to process the entire slide by exploring its local and global information and generating its different synthetic versions. The key idea is to perform the classification and augmentation in feature latent space, reducing the computational cost while preserving the class label of the input. We adopt a deep learning-based multi-classification method and evaluate the contribution given by a conditional generative adversarial network-based data augmentation model on the classifier's performance for three tumor classes in the BRIGHT Challenge dataset. The proposed method has allowed us to achieve an average F1 equal to 69.5, considering only the WSI dataset of the Challenge. The results are comparable to those obtained by the Challenge winning method (71.6), also trained on the annotated tumor region dataset of the Challenge. [ABSTRACT FROM AUTHOR]

Reasoning is a cognitive activity that leverages knowledge to generate solutions to problems. Knowledge representations in the brain require both symbolic and graphical information since visual information is figurative and conveys a large amount of information. Consequently, graphical knowledge representation is often employed in reasoning. Distance-type fuzzy inference utilizes the distance information between the antecedent and the set of facts as the basis for inference. Compared to Mamdani inference, the distance-type fuzzy inference method not only satisfies the convexity and asymptotic properties of the inference results but also adheres to the separation rule (modus ponens), a fundamental principle in inference. This paper discusses extensions of distance-type fuzzy inference methods to handle spatial figures. In this paper, we first explain the distance-type fuzzy inference method. Then, we discuss the concept representation in the feature space and independent parameters that can completely express the characteristics of a figure in space, which are defined as "characteristic parameters". Furthermore, we describe the correspondence between figures and vectors in the feature space, propose a new distance-type fuzzy inference method based on characteristic parameters and describe its characteristics. Finally, an example is used to demonstrate the inference results of this new distance-type fuzzy inference method. [ABSTRACT FROM AUTHOR]

Authorship classification analyzes an author's prior work to identify their writing style, a unique trait of each language and individual author. This research aims to conduct a thorough comparative analysis of various methods for classifying authorship. The study leverages two corpora: AAALitCorpus of Albanian literary texts and CCAT10 of English columns. We evaluate modelgenerated features across different configurations. The richness of the features and the breadth of the analysis provide a significant understanding of the problem, setting a new standard for comprehensive linguistic investigations across multiple languages. The study indicates that machine learning algorithms accurately discern authorial writing styles, highlighting the complexities of classifying authorship in a cross-linguistic context. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Guangdong University of Technology is the property of Journal of Guangdong University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Under the pressure of SDG15.3.1 compliance, it is imperative to solve the land salinization degradation problem in the Yellow River Basin as China’s granary. From the view of geographical scale, six zoning units were divided in the Yellow River Basin with ‘climate-meteorology-geomorphology’ as the main controlling factor, and a salinization inversion model was constructed for each zoning unit. Appropriate surface parameters were selected to construct a three-dimensional feature space according to the individual geographical zones. Based on the cloud data processing capability of the Google Earth Engine platform, a feature space inversion process was applied for automatic inversion of salinization. Salinization distribution maps of the Yellow River Basin in 2015 and 2020 were obtained at 30 m resolution by classifying the salinization inversion result. The distribution and spatiotemporal variation of salinization as well as the causes of salinization were analyzed. Reasonable prevention and control suggestions were subsequently proposed. This study could also be scaled up to larger and more complex geographical regions.

In order to improve the recognition accuracy of upper limb gesture actions, a third-order Butterworth filter was used to denoise surface electromyography ( sEMG) signals and a time sliding window was used to reasonably segment sEMG signals. Integrated EMG value、root mean square value and wavelet packet transform coeffi-cient were used for feature extraction. A feature space method combining time-domain signal with time-frequency signal was proposed, including integrated EMG combined with wavelet packet transform co-efficient ( IEME) and root mean square value combined with wavelet packet transform coefficient ( RMSME) . On the basis of feature space construction, three gesture recognition methods were proposed: support vector machine classifier ( SVM), artificial fish swarm algorithm optimized support vector machine classifier ( AFSA-SVM) and convolutional neural network ( CNN) . A total of 8 upper limb gesture sEMG signals of 10 subjects were collected and compared with the Nina Pro DB2 public dataset. The experimental results show that the feature space IEME is more recognizable than RMSME in both the experimental acquisition data and the Nina Pro DB2 public dataset, and the average recognition accuracy and average training time of the eature space IEME on 1D-CNN are better than that of 2D-CNN. In the experimental acquisition data, the average recognition accuracy of 1D-CNN is as high as 98. 61%, which is 6. 77% and 10. 61% higher than that of SVM and AFSA-SVM, and the average training time of 1D-CNN recognition method is 7. 37 s less than that of SVM and AFSA-SVM by 68. 32 s and 221. 53 s, therefore, the use of feature space IEME and classification model 1D-CNN in gesture sEMG signal recognition classification has advantages. [ABSTRACT FROM AUTHOR]

In unsupervised person re-identification, the traditional asymmetric metric learning alleviates the bias of person images from different views. However, there still exists the issue that the features of the same person are not close to each other in the feature space after asymmetric metric learning. The main reason is that the algorithm cannot overcome other interference except the view, such as the change in person's clothes and scenes, etc. Therefore, the traditional asymmetric metric learning still has the issue of distribution differences in feature space. To address this issue, we propose an asymmetric metric learning method based on distribution regularization constraints for unsupervised person re-identification. First, the JSTL technique pretrains the feature extraction network to obtain robust features. Then, a new asymmetric metric objective function is defined, that is, a distributed regularization constraint term is introduced into the traditional asymmetric metric learning objective function. This method not only alleviates the bias caused by different views, but also effectively overcomes the issue of low recognition accuracy caused by the interference of scenes and clothing changes other than views. Finally, the gradient descent method is used to optimize the objective function, and the optimal metric matrix is obtained by solving the generalized eigenvalue problem. Experiments are implemented on VIPeR, CUHK01, Market-1501, DukeMTMC-Reid, and MSMT17 datasets, and the results show that the Rank-1 values of the algorithm are 35.2 % , 52.62 % , 57.1 % , 42.6 % and 24.5 % . Extensive experimental results show that our algorithm gains significant improvement compared with the state-of-art unsupervised person re-identification algorithms. [ABSTRACT FROM AUTHOR]

The security of a network requires the correct identification and characterization of the attacks through its ports. This involves the follow-up of all the requests for access to the networks by all kinds of users. We consider the frequency of connections and the type of connections to a network, and determine their joint probability. This leads to the problem of determining a joint probability distribution from the knowledge of its marginals in the presence of errors of measurement. Mathematically, this consists of an ill-posed linear problem with convex constraints, which we solved by the method of maximum entropy in the mean. This procedure is flexible enough to accommodate errors in the data in a natural way. Also, the procedure is model-free and, hence, it does not require fitting unknown parameters. [ABSTRACT FROM AUTHOR]

Majority of the learning algorithms used for the training of feedforward neural networks (FNNs), such as backpropagation (BP), conjugate gradient method, etc. rely on the traditional gradient method. Such algorithms have a few drawbacks, including slow convergence, sensitivity to noisy data, local minimum problem, etc. One of the alternatives to overcome such issues is Extreme Learning Machine (ELM), which requires less training time, ensures global optimum and enhanced generalization in neural networks. ELM has a single hidden layer, which poses memory constraints in some problem domains. An extension to ELM, Multilayer ELM (ML-ELM) performs unsupervised learning by utilizing ELM autoencoders and eliminates the need of parameter tuning, enabling better representation learning as it consists of multiple layers. This paper provides a thorough review of ML-ELM architecture development and its variants and applications. The state-of-the-art comparative analysis between ML-ELM and other machine and deep learning classifiers demonstrate the efficacy of ML-ELM in the niche domains of Computer Science which further justifies its competency and effectiveness. [ABSTRACT FROM AUTHOR]

As a new vegetation monitoring index, the KNDVI has certain advantages in characterizing the evolutionary process of regional desertification. However, there are few reports on desertification monitoring based on KNDVI and feature space models. In this study, seven feature parameters, including the kernel normalized difference vegetation index (KNDVI) and Albedo, were introduced to construct different models for desertification remote-sensing monitoring. The optimal desertification remote-sensing monitoring index model was determined with the measured data; then, the spatiotemporal evolution pattern of desertification in Gulang County from 2013 to 2023 was analyzed and revealed. The main conclusions were as follows: (1) Compared with the NDVI and MSAVI, the KNDVI showed more advantages in the characterization of the desertification evolution process. (2) The point–line pattern KNDVI-Albedo remote-sensing index model had the highest monitoring accuracy, reaching 94.93%, while the point–line pattern NDVI-TGSI remote-sensing monitoring index had the lowest accuracy of 54.38%. (3) From 2013 to 2023, the overall desertification situation in Gulang County showed a trend of improvement with a pattern of “firstly aggravation and then alleviation.” Additionally, the gravity center of desertification in Gulang County first shifted to the southeast and then to the northeast, indicating that the northeast’s aggravating rate of desertification was higher than in the southwest during the period. (4) From 2013 to 2023, the area of stable desertification in Gulang County was the largest, followed by the slightly weakened zone, and the most significant transition area was that of extreme desertification to severe desertification. The research results provide important decision support for the precise monitoring and governance of regional desertification.

Members of an online community peer-produce digital artifacts by negotiating different perspectives and personal knowledge bases. These negotiations are manifested in the temporal evolution of the peer-produced artifact. In this study, we conceptualize the evolution of a digital artifact as a trajectory in a feature space. Our theoretical frame suggests that, through negotiations, contributors’ actions “pull” the trajectory and shape its movement in the feature space. We hypothesize that the type of contributors that work on a focal article influences the extent to which that article’s trajectory explores alternative positions within that space, and that the trajectory’s exploration is, in turn, associated with the artifact’s quality. To test these hypotheses, we analyzed the trajectories of wiki articles drawn from two peer-production communities, Wikipedia and Wikia, tracking the evolution of 242 paired articles for over a decade during which the articles went through 536,745 revisions. We found that the contributors who are the most likely to increase the trajectory’s exploration are those that (1) return to work on the focal artifact and (2) are unregistered members in the broader online community. Further, our results show that the trajectory’s exploration has a curvilinear association with article quality, indicating that exploration contributes positively to quality, but that the effect is reversed when exploration exceeds a certain level. The insights derived from this study highlight the value of an artifact-centric approach to increasing our understanding of the dynamics underlying peer-production. [ABSTRACT FROM AUTHOR]

To support pathologists in breast tumor diagnosis, deep learning plays a crucial role in the development of histological whole slide image (WSI) classification methods. However, automatic classification is challenging due to the high-resolution data and the scarcity of representative training data. To tackle these limitations, we propose a deep learning-based breast tumor gigapixel histological image multi-classifier integrated with a high-resolution data augmentation model to process the entire slide by exploring its local and global information and generating its different synthetic versions. The key idea is to perform the classification and augmentation in feature latent space, reducing the computational cost while preserving the class label of the input. We adopt a deep learning-based multi-classification method and evaluate the contribution given by a conditional generative adversarial network-based data augmentation model on the classifier’s performance for three tumor classes in the BRIGHT Challenge dataset. The proposed method has allowed us to achieve an average F1 equal to 69.5, considering only the WSI dataset of the Challenge. The results are comparable to those obtained by the Challenge winning method (71.6), also trained on the annotated tumor region dataset of the Challenge.

Reasoning is a cognitive activity that leverages knowledge to generate solutions to problems. Knowledge representations in the brain require both symbolic and graphical information since visual information is figurative and conveys a large amount of information. Consequently, graphical knowledge representation is often employed in reasoning. Distance-type fuzzy inference utilizes the distance information between the antecedent and the set of facts as the basis for inference. Compared to Mamdani inference, the distance-type fuzzy inference method not only satisfies the convexity and asymptotic properties of the inference results but also adheres to the separation rule (modus ponens), a fundamental principle in inference. This paper discusses extensions of distance-type fuzzy inference methods to handle spatial figures. In this paper, we first explain the distance-type fuzzy inference method. Then, we discuss the concept representation in the feature space and independent parameters that can completely express the characteristics of a figure in space, which are defined as “characteristic parameters”. Furthermore, we describe the correspondence between figures and vectors in the feature space, propose a new distance-type fuzzy inference method based on characteristic parameters and describe its characteristics. Finally, an example is used to demonstrate the inference results of this new distance-type fuzzy inference method.

Random cross-validation (CV) is often used to evaluate geospatial machine learning models, particularly when a limited amount of sample data are available, and collecting an extra test set is unfeasible. However, the prediction locations can be substantially different from the available sample, leading to over-optimistic evaluation results. This has fostered the development of spatial CV methods. Yet these methods only focus on spatial autocorrelation and cannot sufficiently guarantee that the validation subset is a good proxy of the test set with significant differences. In this paper, we propose the spatial+ cross-validation (SP-CV) method. This method, which considers both the geographic and feature spaces, is composed of two stages. The first stage addresses spatial autocorrelation issues by using agglomerative hierarchical clustering to divide the available sample into blocks. The second stage deals with multiple sources of differences. It uses cluster ensembles to split the blocks into training and validation folds based on the locations of the sample data and the values of the covariates and target variable. The proposed method is compared against random and block CV methods in a series of experiments with Amazon basin above ground biomass and California houseprice datasets. Our results show that SP-CV provided the smallest error differences with respect to the reference error. This means that SP-CV produced more representative splits and led to more reliable model evaluations. It suggests that a reliable model evaluation requires to consider both the geographic and the feature spaces in a comprehensive manner.

Desertification is one of the most severe environmental and socioeconomic issues facing the world today. Gonghe Basin is located in the monsoon marginal zone of China, is a sensitive area of climate change in the northeastern of the Qinghai-Tibet Plateau in China, desertification issue has become very severe. Remote sensing monitoring provides an effective technical means for desertification control. In this study, we used Landsat images in 2010 and 2020 to extract desertification information to constructed the Albedo-NDVI feature space in the Gonghe Basin. And then analyzed temporal and spatial evolution of desertification and its driving factors using Geodetector in the Gonghe Basin from 2010 to 2020. The main conclusions are as follows: (1) Albedo-NDVI feature space method can accurately classify desertification information with accuracy of more than 90%, which was benefit to quantitative analysis of desertification. (2) The desertification situation in the Gonghe Basin had improved from 2010 to 2020, especially in the west of the basin, desertification land area decreased by 827.46 km2, and desertification intensity had been obviously reversed. (3) The changes of the desertification in the Gonghe Basin from 2010 to 2020 was affected by both natural and human factors, and the influence of human activities on desertification reversal had increased gradually. The results indicate that the desertification status in the Gonghe Basin had been effectively controlled, and can provide useful basis for the desertification combat in the Gonghe Basin. [ABSTRACT FROM AUTHOR]

Cluster validity indices (CVIs) for evaluating the result of the optimal number of clusters are critical measures in clustering problems. Most CVIs are designed for typical data-type objects called certain data objects. Certain data objects only have a singular value and include no uncertainty, so they are assumed to be information-abundant in the real world. In this study, new CVIs for uncertain data, based on kernel probabilistic distance measures to calculate the distance between two distributions in feature space, are proposed for uncertain clusters with arbitrary shapes, sub-clusters, and noise in objects. By transforming original uncertain data into kernel spaces, the proposed CVI accurately measures the compactness and separability of a cluster for arbitrary cluster shapes and is robust to noise and outliers in a cluster. The proposed CVI was evaluated for diverse types of simulated and real-life uncertain objects, confirming that the proposed validity indexes in feature space outperform the pre-existing ones in the original space. [ABSTRACT FROM AUTHOR]

This article considers a methodological approach to recognize the intrusion of objects into a protected area using an optoelectronic robotic complex, which is based on the existence of a certain characteristic space (a set of signal and geometric features) for each class and type of object. The solved problem of comparing Bayesian a posteriori probabilities of classes (types) of objects is reduced to calculating a priori probabilities and energy distribution functions of signals and geometric parameters of objects, i.e., the likelihood functions of a feature for a specific class (type) of an object. Based on the obtained parameters, the dependences of the probability of correct recognition, omission, false recognition, and confusion of objects on the coefficient of distinctness when the object's signal energy deviates from the standard energy are analyzed. The results obtained are required to solve the problem of the adaptive group control of robotic complexes when solving operational and tactical and tasks in an uncertain dynamic environment. [ABSTRACT FROM AUTHOR]

Few studies of monitoring rocky desertification based on Sentinel-2 MSI image with higher spatial and temporal resolutions have been reported. In this paper, nine typical surface parameters that derived from Sentinel-2 MSI image including three red edge indices have been introduced to establish two types of feature space monitoring index: point-to-line type and point-to-point type. The results show that: (1) The feature of space monitoring index based on Sentinel-2 MSI image has a higher applicability in obtaining rocky desertification information for karst zones; (2) The point-to-line BSI (bare soil index)-NDre1 (red edge index 1)feature space monitoring index of rocky desertification has the best applicability with R2=95.1% and root mean square error (RMSE)=0.872; (3) The feature space monitoring indexes that composed of red edge indices have a higher inversion accuracy, which indicates that the red edge indices are sensitive indicators of the rocky desertification condition and process;(4) The intensive and severe rocky desertification zones are mostly located in the southeast and central parts. The 3-dimensional or multidimensional feature space model will be constructed to improve the inversion accuracy of rocky desertification in further studies. The research results can provide methodological supports for revealing the rocky desertification evolution laws.

Accurate monitoring of soil moisture and the development of timely interventions are important to reduce the social and economic losses caused by drought. Compared to short-wave infrared (SWIR) and thermal infrared (TIR), near-infrared (NIR) and visible bands are widely used in almost all optical satellites. Drought monitoring using NIR and visible bands is therefore more relevant for optical satellites. Among the visible bands, the red band is often used in combination with the NIR band for drought monitoring due to its sensitivity to vegetation. However, current drought indexes based on the NIR and the red band applied to areas of high vegetation suffer from insufficient accuracy or tedious calculations. In this study, the ratio drought index (RDI) was developed after constructing a new feature space by examining the spectral properties of soil and vegetation at different water levels in the NIR and red bands. The accuracy of soil moisture inversion under two types of bare soil and vegetation was evaluated using in situ data from Tai’an City, Shandong Province. The perpendicular drought index (PDI) and modified perpendicular drought index (MPDI) were also used to compare for the RDI. The results showed that the RDI correlation coefficients (R2) of 0.653 and 0.641 were better than the MPDI of 0.616 and 0.594 and the PDI of 0.602 and 0.546 for soil moisture measurements from vegetation and bare soil cover. The RDI attenuates the effect of vegetation on soil moisture inversion, as its root mean square error (RMSE) in vegetated areas is lower than that of the PDI and MPDI. The RDI calculation can be used as a theoretical guide for large-scale soil moisture estimation because it is fast, accurate and does not require additional quantitative remote sensing inversion factors. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Frontiers of Computer Science & Technology is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Summary: Privacy leakage in images attracts increasing concerns these days, as photos uploaded to large social platforms are usually not processed by proper privacy protection mechanisms. Moreover, with advanced artificial intelligence (AI) tools such as deep neural network (DNN), an adversary can detect people's identities and collect other sensitive personal information from images at an unprecedented scale. In this paper, we introduce a novel face image de‐identification framework using adversarial perturbations in the feature space. Manipulating the feature space vector ensures the good transferability of our framework. Moreover, the proposed feature space adversarial perturbation generation algorithm can successfully protect the identity‐related information while ensuring the other attributes remain similar. Finally, we conduct extensive experiments on two face image datasets to evaluate the performance of the proposed method. Our results show that the proposed method can generate real‐looking privacy‐preserving images efficiently. Although our framework has only been tested on two real‐life face image datasets, it can be easily extended to other types of images. [ABSTRACT FROM AUTHOR]

The Dynamic World product is a discrete land cover classification of Sentinel 2 reflectance imagery that is global in extent, retrospective to 2015, and updated continuously in near real time. The classifier is trained on a stratified random sample of 20,000 hand-labeled 5 × 5 km Sentinel 2 tiles spanning 14 biomes globally. Since the training data are based on visual interpretation of image composites by both expert and non-expert annotators, without explicit spectral properties specified in the class definitions, the spectral characteristics of the classes are not obvious. The objective of this study is to quantify the physical distinctions among the land cover classes by characterizing the spectral properties of the range of reflectance present within each of the Dynamic World classes over a variety of landscapes. This is achieved by comparing both the eight-class probability feature space (excluding snow) and the maximum probability class assignment (label) distributions to continuous land cover fraction estimates derived from a globally standardized spectral mixture model. Standardized substrate, vegetation, and dark (SVD) endmembers are used to unmix nine Sentinel 2 reflectance tiles from nine spectral diversity hotspots for comparison between the SVD land cover fraction continua and the Dynamic World class probability continua and class assignments. The variance partition for the class probability feature spaces indicates that eight of these nine hotspots are effectively five-dimensional to 95% of variance. Class probability feature spaces of the hotspots all show a tetrahedral form with probability continua spanning multiple classes. Comparison of SVD land cover fraction distributions with maximum probability class assignments (labels) and probability feature space distributions reveal a clear distinction between (1) physically and spectrally heterogeneous biomes characterized by continuous gradations in vegetation density, substrate albedo, and structural shadow fractions, and (2) more homogeneous biomes characterized by closed canopy vegetation (forest) or negligible vegetation cover (e.g., desert, water). Due to the ubiquity of spectrally heterogeneous biomes worldwide, the class probability feature space adds considerable value to the Dynamic World maximum probability class labels by offering users the opportunity to depict inherently gradational heterogeneous landscapes otherwise not generally offered with other discrete thematic classifications. [ABSTRACT FROM AUTHOR]

With view to applications to harmonic and stochastic analysis of infinite network/graph models, we introduce new tools for realizations and transforms of positive definite kernels (p.d.) $ K $ and their associated reproducing kernel Hilbert spaces. With this we establish two kinds of factorizations: (i) Probabilistic: Starting with a positive definite kernel $ K $ we analyze associated Gaussian processes $ V $. Properties of the Gaussian processes will be derived from certain factorizations of $ K $, arising as a covariance kernel of $ V $. (ii) Geometric analysis: We discuss families of measure spaces arising as boundaries for $ K $. Our results entail an analysis of a partial order on families of p.d. kernels, a duality for operators and frames, optimization, Karhunen–Loève expansions, and factorizations. Applications include a new boundary analysis for the Drury-Arveson kernel, and for certain fractals arising as iterated function systems; and an identification of optimal feature spaces in machine learning models. [ABSTRACT FROM AUTHOR]

Surface soil moisture (SSM) is an important variable for controlling the energy exchange between earth's surface and atmosphere. Accurate estimation of the distribution of SSM and its dynamic changes is increasingly essential in the study of hydrology, ecology, agriculture, geological hazard monitoring, and global climate change. This paper presents a new model for SSM retrieval (named, KSSM) through assimilating soil porosity with a two-dimensional feature space composed of land surface temperature and vegetation index (LST-VI). The KSSM model takes into account the physical properties of soil porosity, so it is suitable for SSM retrieval of various soil types in different regions. The KSSM model was validated using the measured data, and a comparison analysis with other data was conducted. Through the sensitivity analysis, it is concluded that the error of SSM retrieved by the KSSM model is less than 0.04 m3 m−3 and the KSSM model is robust. High accuracy of volumetric SSM estimation, error within 10% compared with the measured data, can be achieved by the KSSM model without the need for an in-situ SSM. [ABSTRACT FROM AUTHOR]

Under the pressure of SDG15.3.1 compliance, it is imperative to solve the land salinization degradation problem in the Yellow River Basin as China's granary. From the view of geographical scale, six zoning units were divided in the Yellow River Basin with 'climate-meteorology-geomorphology' as the main controlling factor, and a salinization inversion model was constructed for each zoning unit. Appropriate surface parameters were selected to construct a three-dimensional feature space according to the individual geographical zones. Based on the cloud data processing capability of the Google Earth Engine platform, a feature space inversion process was applied for automatic inversion of salinization. Salinization distribution maps of the Yellow River Basin in 2015 and 2020 were obtained at 30 m resolution by classifying the salinization inversion result. The distribution and spatiotemporal variation of salinization as well as the causes of salinization were analyzed. Reasonable prevention and control suggestions were subsequently proposed. This study could also be scaled up to larger and more complex geographical regions. [ABSTRACT FROM AUTHOR]

Ordinalysis is a software that enables dimension reduction, visualization and quantitative ordinality analysis of ordinal data. It is provided as a standalone executable file with a video tutorial. Applications of the software are shown on an ordinal synthetic dataset, ordinal real feature spaces and an ordinal image dataset. Ordinalysis allows the detection of violations of ordinality, the selection of the dimension reduction to preserve ordinality best or the demonstration of existing semantic ordinality in a set of raw data.

The security of a network requires the correct identification and characterization of the attacks through its ports. This involves the follow-up of all the requests for access to the networks by all kinds of users. We consider the frequency of connections and the type of connections to a network, and determine their joint probability. This leads to the problem of determining a joint probability distribution from the knowledge of its marginals in the presence of errors of measurement. Mathematically, this consists of an ill-posed linear problem with convex constraints, which we solved by the method of maximum entropy in the mean. This procedure is flexible enough to accommodate errors in the data in a natural way. Also, the procedure is model-free and, hence, it does not require fitting unknown parameters.

Copyright of Remote Sensing for Natural Resources is the property of Remote Sensing for Natural Resources Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Neural Networks are well known for its performance to classify and cluster data sets via multiple layers of networks passing and transforming information pictured by raw data. The feature layer projects the raw data into a space spanned by hidden features. To understand data representations in both original (i.e., image) and feature spaces, the main purpose of this research is to analyze the clustering performance with different feature representations. Naturally, distance measures have a great impact on clustering performance. Different distances and their combinations are tested on both the original and feature spaces. The combined distances were obtained by using different optimal weights that minimize classification errors in different measures via a series of optimization models. These weights were multiplied by their respective distances in order to create the combined distance. Clustering was evaluated using silhouette scores. The feature space in general has better performance, in terms of clustering, than the image space, with Cosine Similarity being the best distance for both the image space and feature space. [ABSTRACT FROM AUTHOR]

Objective: Sonomyography, or ultrasound-based sensing of muscle deformation, is an emerging modality for upper limb prosthesis control. Although prior studies have shown that individuals with upper limb loss can achieve successful motion classification with sonomyography, it is important to better understand the time-course over which proficiency develops. In this study, we characterized user performance during their initial and subsequent exposures to sonomyography. Method: Ultrasound images corresponding to a series of hand gestures were collected from individuals with transradial limb loss under three scenarios: during their initial exposure to sonomyography (Experiment 1), during a subsequent exposure to sonomyography where they were provided biofeedback as part of a training protocol (Experiment 2), and during testing sessions held on different days (Experiment 3). User performance was characterized by offline classification accuracy, as well as metrics describing the consistency and separability of the sonomyography signal patterns in feature space. Results: Classification accuracy was high during initial exposure to sonomyography (96.2 ± 5.9%) and did not systematically change with the provision of biofeedback or on different days. Despite this stable classification performance, some of the feature space metrics changed. Conclusions: User performance was strong upon their initial exposure to sonomyography and did not improve with subsequent exposure. Clinical Impact: Prosthetists may be able to quickly assess if a patient will be successful with sonomyography without submitting them to an extensive training protocol, leading to earlier socket fabrication and delivery.

The generative intelligent design of building structures has been a rapidly evolving field in recent years, with shear wall structures efficiently designed using generative adversarial networks (GANs). Nevertheless, using RGB images to represent structural design drawings may not accurately capture the correlations and distinctions between building components, leading to shear-wall structure design inaccuracies. To address these challenges, this study proposes an optimized data representation and understanding method for the intelligent design of shear wall structures. Specifically, a component-based feature space data representation method was introduced to achieve a more accurate description of drawing features. A feature mask was also employed to precisely locate the key target areas for generating shear walls, enhancing the model's understanding of the data features. Additionally, a GAN with global and local discriminators was established to enhance the neural network's ability to generate shear wall designs at both the global image and local component levels. Finally, a training method for a GAN with dual discriminators based on feature masks is proposed. Experimental analyses and case studies showed that, compared with existing widely used GAN-based designs, the structural designs generated by the GAN trained with the proposed data representation and understanding methods demonstrated superior consistency with engineer-designed structures and outperformed structural seismic resistance. • Optimized data representation and understanding method for structural design. • Improve the model's understanding of component relationships in drawings. • Enhance performance on data with sparse features. • Improve the model's ability for joint learning of both global and local design. [ABSTRACT FROM AUTHOR]

In terms of crowd counting methods, usually a large number of labeled samples were meeded to train a counting model. In practical applications, to solve the problem of high cost and low efficiency of manually labeling samples, a new sample labeling method for crowd counting was proposed to obtain sufficient training samples. Considering the fact that the distribution of video images having the identical or analogous crowd counts is close to each other in the feature space, with the first neighbor clustering, the video images with similar distance in feature space were divided into the same clusters. A small number of key samples with diversity and representativeness from each cluster were selected by using active sampling strategy and then labeled manually. In the end, the number of labeled samples were propagated to the rest unlabeled samples in each cluster to realize the rapid labeling of samples. The results show that the proposed labeling method can effectively reduce the labeling burden in the crowd counting task, which can achieve better prediction accuracy only with 30% of the labeled samples from other crowd counting methods. Besides, the counting models trained by the obtained sample labels have consistent prediction results compared to the counting models trained by truth value.