Improvement In Overall Accuracy Research Articles

The Impact of Multiple Pond Conditions on the Performance of Dike-Pond Extraction

Dike-ponds in fisheries often present multiple pond conditions such as pure, suspended sediment, water bloom, semidry conditions, etc. However, the impact of these conditions on the performance of extracting dike-pond from remote sensing images has not been studied. To solve this problem, we explore the existence of such impacts by comparing the performance of four rule-based methods in two groups of test regions. The first group has few multiple pond conditions, while the second has more. The results show that various measure values deteriorate as the proportion of multiple pond conditions in the regions increases. All four methods performed worse in the second group than the first, where the overall accuracy decreased by 8.80%, misclassification error increased by 3.69%, omission error raised by 10.53%, and correct quantity rate dropped by 8.23%, respectively. The extraction method that ingested multiple pond conditions performed indistinguishably from the other methods in the first group. However, it outperformed the other methods in the second group, with a 4.22% improvement in overall accuracy, a 10.25% decrease in misclassification error, and a 19.03% increase in the correct quantity rate. These findings suggest that multiple pond conditions can negatively impact the extraction performance and should be considered in dike-pond applications that require a precise pond size, number, and shape.

A supervised deep neural network approach with standardized targets for enhanced accuracy of IVIM parameter estimation from multi-SNR images.

Extraction of intravoxel incoherent motion (IVIM) parameters from noisy diffusion‐weighted (DW) images using a biexponential fitting model is computationally challenging, and the reliability of the estimated perfusion‐related quantities represents a limitation of this technique. Artificial intelligence can overcome the current limitations and be a suitable solution to advance use of this technique in both preclinical and clinical settings. The purpose of this work was to develop a deep neural network (DNN) approach, trained on numerical simulated phantoms with different signal to noise ratios (SNRs), to improve IVIM parameter estimation. The proposed approach is based on a supervised fully connected DNN having 3 hidden layers, 18 inputs and 3 targets with standardized values. 14 × 103 simulated DW images, based on a Shepp–Logan phantom, were randomly generated with varying SNRs (ranging from 10 to 100). 7 × 103 images (1000 for each SNR) were used for training. Performance accuracy was assessed in simulated images and the proposed approach was compared with the state‐of‐the‐art Bayesian approach and other DNN algorithms. The DNN approach was also evaluated in vivo on a high‐field MRI preclinical scanner. Our DNN approach showed an overall improvement in accuracy when compared with the Bayesian approach and other DNN methods in most of the simulated conditions. The in vivo results demonstrated the feasibility of the proposed approach in real settings and generated quantitative results comparable to those obtained using the Bayesian and unsupervised approaches, especially for D and f, and with lower variability in homogeneous regions. The DNN architecture proposed in this work outlines two innovative features as compared with other studies: (1) the use of standardized targets to improve the estimation of parameters, and (2) the implementation of a single DNN to enhance the IVIM fitting at different SNRs, providing a valuable alternative tool to compute IVIM parameters in conditions of high background noise.

Conjunctive Use of Landsat-8 OLI and MODIS Data for Delineation of Burned Areas

For regional-level monitoring of burned areas, Moderate Resolution Imaging Spectroradiometer (MODIS ) MCD64A1 and MCD45A1 products have been operationally used. However, because of their coarser spatial resolution, such products do not allow for detection of small patches (<50 ha) of burned areas, which are very important for modeling gas emissions. In order to bridge this gap, we undertook a study to evaluate the synergy of MCD64A1 and Landsat-8 Operational Land Imager (OLI ) data for delineating burned areas in part of the mountainous terrain of the Himalayas, northern India. The approach involves generating a differenced normalized burn ratio image from Landsat-8 OLI data before and after fire events, refining the burned areas thus delineated using the MCD64A1 burned-area product and field data, and estimating accuracy. An overall mapping accuracy of 90.0% was achieved using Landsat-8 OLI data. Inclusion of the MODIS MCD64A1 burned-area product resulted in marginal improvement in overall accuracy (to 94.0%).

Effector-GAN: prediction of fungal effector proteins based on pretrained deep representation learning methods and generative adversarial networks.

Phytopathogenic fungi secrete effector proteins to subvert host defenses and facilitate infection. Systematic analysis and prediction of candidate fungal effector proteins are crucial for experimental validation and biological control of plant disease. However, two problems are still considered intractable to be solved in fungal effector prediction: one is the high-level diversity in effector sequences that increases the difficulty of protein feature learning, and the other is the class imbalance between effector and non-effector samples in the training dataset. In our study, pretrained deep representation learning methods are presented to represent multiple characteristics of sequences for predicting fungal effectors and generative adversarial networks are adapted to create synthetic feature samples to address the data imbalance problem. Compared with the state-of-the-art fungal effector prediction methods, Effector-GAN shows an overall improvement in accuracy in the independent test set. Effector-GAN offers a user-friendly interface to inspect potential fungal effector proteins (http://lab.malab.cn/~wys/webserver/Effector-GAN). The Python script can be downloaded from http://lab.malab.cn/~wys/gitlab/effector-gan. Supplementary data are available at Bioinformatics online.

Automated assessment of subjective assignments: A hybrid approach

Machine learning (ML) has recently gained popularity in a variety of domains for automating tasks. This has also been used in the evaluation of student responses/ assignments. This paper presents a qualitatively enhanced methodology for automated score prediction of subjective assignments. To consider all the major aspects of a human grader, 21 linguistic features related to syntactical, grammatical, sentimental, and readability are analyzed from the assignments. This study makes use of the ASAP competition dataset from Kaggle. The evaluation metric used is Quadratic Weighted Kappa (QWK), which measures the agreement between the human graded score and the predicted score. The effect of appropriate feature selection has been observed using Mutual Information Regression. Four ML algorithms are investigated on identified linguistic features. 3 Layer Neural Network with feature selection performed well among all chosen ML algorithms with average QWK of 0.678. To include the benefits of deep learning, a new hybrid model (LF-BiLSTM-att-FS) is proposed that combines a higher level deep neural network (DNN) with the selected features. Pre-trained Glove embedding is used to include contextual information of the text. The proposed model results demonstrated an improvement in overall accuracy, with an average QWK value of 0.768.

Assessing Online Symptom Checker Accuracy in Diagnosing Men's Health Conditions

ABSTRACT Introduction Many medical sites offer algorithm-based tools where patients can enter symptoms to receive potential diagnoses. The efficacy of these online symptom checkers (OSCs) in properly diagnosing male sexual health issues has not been explored. Objective We analyzed the accuracy of the four most commonly used OSCs in diagnosing erectile dysfunction (ED), scrotal pain (SP), Peyronie's disease (PD), and low testosterone (LT). Methods One-hundred and sixty artificial vignettes were created by de-identifying recent initial outpatient consults presenting to discuss ED (40), SP (40), PD (40), and LT (40). ED and PD were selected for assessment because they were straightforward diagnosis. In contrast, SP and LT was selected because of varied etiologies and clinical presentations. The vignettes were entered into the 4 most frequently used OSCs as determined by web traffic analysis tools. The top 4 OSCs in order of use are WebMD (WMD), MedicineNet (MN), Everyday Health (EH) and Sutter Health (SH). The top 5 conditions listed in the OSC differential diagnosis were recorded and scored. If less than 5 conditions were offered, the available conditions were recorded and scored. The primary outcome was the accuracy of primary (Top 1) condition listed. The secondary outcome was frequency of the correct diagnosis appearing in the Top 3, or 5 of the suggested conditions. Results WMD did not offer the correct diagnosis for any ED or PD regardless of the number of conditions included from the differential vignettes (Table 1). However, WMD did successfully diagnose SP and LT in 22.5% and 95% of patients, respectively. Comparatively, EH was only able to diagnose SP 20% of the time, and failed to diagnose ED, PD, or LT on all occasions. MN successfully diagnosed ED, PD, SP, and LT in 100%, 98%, 27.5%, and 0% of patients, respectively. On average, MN was the most accurate OSC when including Top 1 (48.25%), Top 3 (53.88%), and Top 5 (56%) suggested conditions. Lastly, SH saw the largest overall improvement in accuracy when using the Top 5 suggested conditions compared to Top 1 (50% vs. 18%). SH correctly diagnosed ED, SP, and LT in 100%, 20%, and 80% of patients, respectively. The OSCs were most accurate diagnosing ED and least accurate diagnosing SP when using the Top 1 (37.5% vs. 6.88%), Top 3 (47.5% vs. 17.5%) and Top 5 (50% vs. 24.5%) of the suggested conditions. Conclusions WebMD, which receives nearly five times the web traffic of the next leading online symptom checker, and Everyday Health were both incapable of diagnosing ED and PD in this series. Only MedicineNet offered diagnoses for PD. All OSCs were routinely poor at suggesting precise diagnoses for scrotal pain, even when including 5 potential options in the differential. LT was only diagnosed by WMD and SH. Patients should be cautioned regarding the accuracy of OSCs. Disclosure Any of the authors act as a consultant, employee or shareholder of an industry for: Coloplast

IAGC: Interactive Attention Graph Convolution Network for Semantic Segmentation of Point Clouds in Building Indoor Environment

Point-based networks have been widely used in the semantic segmentation of point clouds owing to the powerful 3D convolution neural network (CNN) baseline. Most of the current methods resort to intermediate regular representations for reorganizing the structure of point clouds for 3D CNN networks, but they may neglect the inherent contextual information. In our work, we focus on capturing discriminative features with the interactive attention mechanism and propose a novel method consisting of the regional simplified dual attention network and global graph convolution network. Firstly, we cluster homogeneous points into superpoints and construct a superpoint graph to effectively reduce the computation complexity and greatly maintain spatial topological relations among superpoints. Secondly, we integrate cross-position attention and cross-channel attention into a single head attention module and design a novel interactive attention gating (IAG)-based multilayer perceptron (MLP) network (IAG–MLP), which is utilized for the expansion of the receptive field and augmentation of discriminative features in local embeddings. Afterwards, the combination of stacked IAG–MLP blocks and the global graph convolution network, called IAGC, is proposed to learn high-dimensional local features in superpoints and progressively update these local embeddings with the recurrent neural network (RNN) network. Our proposed framework is evaluated on three indoor open benchmarks, and the 6-fold cross-validation results of the S3DIS dataset show that the local IAG–MLP network brings about 1% and 6.1% improvement in overall accuracy (OA) and mean class intersection-over-union (mIoU), respectively, compared with the PointNet local network. Furthermore, our IAGC network outperforms other CNN-based approaches in the ScanNet V2 dataset by at least 7.9% in mIoU. The experimental results indicate that the proposed method can better capture contextual information and achieve competitive overall performance in the semantic segmentation task.

Summary accuracy feedback and the left digit effect in number line estimation.

A robust left digit effect arises in number line estimation such that adults’ estimates for numerals with different hundreds place digits but nearly identical magnitudes are systematically different from one another (e.g., 299 is placed too far to the left of 302). In two experiments, we investigate whether brief feedback interventions designed to increase task effort can reduce or eliminate the left digit effect in a self-paced 0–1,000 number line estimation task. Participants were assigned to complete three blocks of 120 trials each where the middle block contained feedback or no feedback. Feedback was in the form of summary accuracy scores (Experiment 1; N = 153) or competitive (summary) accuracy scores (Experiment 2; N = 145). In both experiments, planned analyses revealed large left digit effects in all blocks regardless of feedback condition. Feedback did not lead to a reduction in the left digit effect in either experiment, but improvements in overall accuracy were observed. We conclude that there are no changes in the left digit effect resulting from either summary accuracy feedback or competitive accuracy feedback. Also reported are exploratory analyses of trial characteristics (e.g., whether 299 is presented before or after 302) and the left digit effect.Supplementary InformationThe online version contains supplementary material available at 10.3758/s13421-022-01278-2.

Applying multidate Sentinel-2 data for forest-type classification in complex broadleaf forest stands

AbstractBiodiversity assessment and forest management require accurate tree species maps, which can be provided by remote sensing. Whereas the application of high-spatial resolution remote sensing data is constrained by high costs, Sentinel-2 (S2) satellites provide free imagery with appropriate spatial, spectral and temporal resolutions for mapping of various forest traits across larger spatial scales. Here we assessed the potential of multidate S2 as well as a Digital Elevation Model (DEM) in classifying tree species across a highly structured and heterogeneous broadleaf forest ecosystem in the Hyrcanian zone of northern Iran. We applied multidate S2 and DEM data as input to a variable selection using random forests algorithm for feature reduction. Ten forest types were classified using random forest algorithm and to evaluate the results we computed area-adjusted confusion matrices. Classifications based on single-date S2 data reached overall accuracies of 67–74 per cent, whereas results for multidate S2 images increased the accuracy by ~28 per cent. Joint use of DEM data along with multidate S2 images showed improvement of overall accuracy by ~3 per cent. In addition, we studied the effect of topographic correction of S2 data on classification performance. The results imply that applying topographically corrected imagery had no significant effect on the classification accuracy. Our results demonstrate the high potential of freely available multisource remotely sensed data for broadleaf tree species classification across complex broad-leaved forest landscapes.

Cross-Domain Classification of Multisource Remote Sensing Data Using Fractional Fusion and Spatial-Spectral Domain Adaptation

Limitation of labeled samples has always been a challenge for hyperspectral image (HSI) classification. In real remote sensing applications, we encounter a situation where an HSI scene is not labeled at all. To solve this problem, cross-domain learning methods are developed by utilizing another HSI scene with similar land covers and sufficient labeled samples. However, the disparity between HSI scenes is still a challenge in reducing the classification performance, which may be affected by variations in illumination and weather. As a robust supplement to these variations, light detection and ranging (LiDAR) data provides stable elevation and spatial information. In this paper, we propose a multisource cross-domain classification method using fractional fusion and spatial-spectral domain adaptation to reduce the disparity between scenes. The spatial information of HSI is preserved by fractional differential masks (FrDM) first. Then the LiDAR data is utilized for spectral alignment of HSI. The utilization of LiDAR data reduces the pixel-level disparity between scenes. At last, a spatial-spectral domain adaptation network (SSDA) is proposed to reduce domain shift at the feature level and extract discriminative spatial-spectral features. Experimental results on HSI and LiDAR scenes show <inline-formula><tex-math notation="LaTeX">$5\%-10\%$</tex-math></inline-formula> improvements in overall accuracy compared with state-of-the-art methods.

The impacts of vegetation indices from UAV-based RGB imagery on land cover classification using ensemble learning

The production of land use and land cover (LULC) maps using UAV images obtained by RGB cameras offering very high spatial resolution has recently increased. Vegetation indices (VIs) have been widely used as an important ancillary data to increase the limited spectral information of the UAV image in pixel-based classification. The main goal of this study is to analyze the effect of frequently used RGB-based VIs including green leaf index (GLI), red- green-blue vegetation index (RGBVI) and triangular greenness index (TGI) on the classification of UAV images. For this purpose, five different dataset combinations comprising of RGB bands and VIs were formed. In order to evaluate their effects on thematic map accuracy, four ensemble learning methods, namely RF, XGBoost, LightGBM and CatBoost were utilized in classification process. Classification results showed that the use of RGB UAV image with VIs increased the overall accuracy (OA) values in all cases. On the other hand, the highest OA values were calculated with the use of Dataset-5 (i.e. RGB bands and all VIs considered). Additionally, the classification result of Dataset-4 (i.e. RGB bands and TGI) showed superior performance compared to Dataset-2 (i.e. RGB bands and GLI) and Dataset-3 (i.e. RGB bands and RGBVI). All in all, the TGI was found to be useful for improving classification accuracy of UAV image having limited spectral information compared to GLI and RGBVI. The improvement in overall accuracy reached to 2% with the use of RGB bands and TGI index. Furthermore, within the ensemble algorithms, CatBoost produced the highest overall accuracy (92.24%) with the dataset consist of RBG bands and all VIs considered.

Generating 2m fine-scale urban tree cover product over 34 metropolises in China based on deep context-aware sub-pixel mapping network

Contrast to the global forest, few trees live in cities but contribute significantly to urban environment and human health. However, the classical satellite-derived land cover/forest cover products with limited resolution are not fine enough for the identification of urban tree, which is usually appeared in small size and intersected with infrastructure. To relieve the dilemma, this study developed an urban tree specific sub-pixel mapping (SPM) architecture with deep learning approach, which aimed to generate 2m fine-scale urban tree cover product from 10 m Sentinel-2 images for large-scale area of 34 metropolises in China. The proposed approach has remarkable reconstruction ability for delineating the contextual characteristic of the urban tree patterns, and reliable generalization ability to large-scale area. In addition, this study creates a large-volume urban tree cover dataset (UTCD) with 0.13 billion urban tree samples at 2 m resolution, which fills the deficiency of standard dataset in urban tree cover research field. Quantitative analysis of our products was conducted on two typical study sites of Beijing and Wuhan. The results show that our products recover averagely more than 58.72% of urban tree covers that have been underestimated in the existing land cover/forest cover products, and outperforms the state-of-the-art approach both visually and quantitatively, by averagely 11.31% improvement in overall accuracy. From our annual products during 2016–2020, we found an evolution characteristic of urban tree cover: it is more stable in developed cities like Beijing, while more fluctuated in developing cities like Wuhan, and the alteration are usually concentrated at the outer-ring of downtown, which may be caused by the municipal planning and the land development of real estate industry.

Synergistic Use of Geospatial Data for Water Body Extraction from Sentinel-1 Images for Operational Flood Monitoring across Southeast Asia Using Deep Neural Networks

Deep learning is a promising method for image classification, including satellite images acquired by various sensors. However, the synergistic use of geospatial data for water body extraction from Sentinel-1 data using deep learning and the applicability of existing deep learning models have not been thoroughly tested for operational flood monitoring. Here, we present a novel water body extraction model based on a deep neural network that exploits Sentinel-1 data and flood-related geospatial datasets. For the model, the U-Net was customised and optimised to utilise Sentinel-1 data and other flood-related geospatial data, including digital elevation model (DEM), Slope, Aspect, Profile Curvature (PC), Topographic Wetness Index (TWI), Terrain Ruggedness Index (TRI), and Buffer for the Southeast Asia region. Testing and validation of the water body extraction model was applied to three Sentinel-1 images for Vietnam, Myanmar, and Bangladesh. By segmenting 384 Sentinel-1 images, model performance and segmentation accuracy for all of the 128 cases that the combination of stacked layers had determined were evaluated following the types of combined input layers. Of the 128 cases, 31 cases showed improvement in Overall Accuracy (OA), and 19 cases showed improvement in both averaged intersection over union (IOU) and F1 score for the three Sentinel-1 images segmented for water body extraction. The averaged OA, IOU, and F1 scores of the ‘Sentinel-1 VV’ band are 95.77, 80.35, and 88.85, respectively, whereas those of ‘band combination VV, Slope, PC, and TRI’ are 96.73, 85.42, and 92.08, showing improvement by exploiting geospatial data. Such improvement was further verified with water body extraction results for the Chindwin river basin, and quantitative analysis of ‘band combination VV, Slope, PC, and TRI’ showed an improvement of the F1 score by 7.68 percent compared to the segmentation output of the ‘Sentinel-1 VV’ band. Through this research, it was demonstrated that the accuracy of deep learning-based water body extraction from Sentinel-1 images can be improved up to 7.68 percent by employing geospatial data. To the best of our knowledge, this is the first work of research that demonstrates the synergistic use of geospatial data in deep learning-based water body extraction over wide areas. It is anticipated that the results of this research could be a valuable reference when deep neural networks are applied for satellite image segmentation for operational flood monitoring and when geospatial layers are employed to improve the accuracy of deep learning-based image segmentation.

PrPred-DRLF: Plant R protein predictor using deep representation learning features.

Plant resistance (R) proteins play a significant role in the detection of pathogen invasion. Accurately predicting plant R proteins is a key task in phytopathology. Most plant R protein predictors are dependent on traditional feature extraction methods. Recently, deep representation learning methods have been successfully applied in solving protein classification problems. Motivated by this, we propose a new computational approach, called prPred-DRLF, which uses deep representation learning feature models to encode the amino acids as numerical vectors. The results show that the fused features of bidirectional long short-term memory (BiLSTM) embedding and unified representation (UniRep) embedding have a better performance than other features for plant R protein identification using a light gradient boosting machine (LGBM) classifier. The model was evaluated using an independent test achieving an accuracy of 0.956, F1-score of 0.933, and area under the receiver operating characteristic (ROC) curve (AUC) of 0.997. Meanwhile, compared with the state-of-the-art prPred and HMMER method, prPred-DRLF shows an overall improvement in accuracy, F1-score, AUC, and recall. prPred-DRLF is a higher-performance plant R protein prediction tool based on two kinds of deep representation learning technologies and offers a user-friendly interface for inspecting possible plant R proteins. We hope that prPred-DRLF will become a useful tool for biological research. A user-friendly webserver for prPred-DRLF is freely accessible at http://lab.malab.cn/soft/prPred-DRLF. The Python script can be downloaded from https://github.com/Wangys-prog/prPred-DRLF.

Individual and collaborative processing of written corrective feedback affects second language writing accuracy and revision

Although substantial evidence exists showing written corrective feedback (WCF) helps second language (L2) learners improve their writing accuracy, there is a paucity of research that has examined L2 writing accuracy in connection with both revision behavior and processing of WCF. Using a quasi-experimental research design, 44 learners from four Academic and Professional Writing courses were divided into two groups: individual processing (n = 20) and collaborative processing (n = 24). The participants completed four timed writing tasks and received indirect WCF from a course teacher that used different error codes to target nine error types. The participants processed the WCF individually or collaboratively before revising. Participants’ revision behavior (the number of errors successfully resolved, unresolved, and abandoned) and writing accuracy (total number of errors and total number of errors by error type) were analyzed and compared. The results showed the collaborative processing group’s performance in terms of resolved, unresolved, and abandoned errors was superior to the individual processing group. While both groups exhibited improvements in overall accuracy, the collaborative processing group produced fewer verb and word choice errors after collaboratively processing three rounds of WCF. The results are discussed in terms of the importance of L2 writing teachers’ pedagogical practices.

Cannabis detection from optical and RADAR data fusion: A comparative analysis of the SMILE machine learning algorithms in Google Earth Engine

Accurate crop mapping for agricultural monitoring requires the use of robust image classification algorithms. Cultivation of illegal cannabis is common in the Bekaa plain, Lebanon, as in some other areas of the world (Morocco, Afghanistan). Identification of cultivation sites is money and time consuming since it highly relies on knowledge of cultivation areas and field surveys. Machine-learning-based image classification provides an alternative method for cannabis detection. This paper presents a comparative analysis of four machine-learning classifiers implemented in Google Earth Engine: Random Forest (RF), Gradient Boosting (GBT), Classification and Regression Tree (CART), and Support Vector Machine (SVM) for cannabis and other crop type classification. We implement and test several image fusion approaches for optical (Sentinel-2, Landsat) and Synthetic Aperture Radar (SAR) imagery (Sentinel-1) along with several combinations of surface textural features from Sentinel-1. Six different crop groups were classified over three years (2016, 2017, and 2018) in the Bekaa valley of Lebanon. In general, although SVM outperformed RF, GBT, and CART classifiers with an overall classification accuracy of more than 90%, RF and GBT provided more consistent results in terms of the cannabis cultivated area. SVM appears to be sensitive to the size of the training data. RF and GBT estimate an average cannabis area of 2800 ha in 2016, 2983 ha in 2017, and 5900 ha in 2018. Our results demonstrate that the fusion of radar and optical imagery can improve image classification accuracy by up to 5%. A marginal improvement in overall accuracy was observed when textural features were added. This is the first study that used image fusion and ML to estimate illegal cannabis cultivation areas in Lebanon and help evaluate the contribution of cannabis to the local economy. The combination of machine learning algorithms and imagery fusion proved reliable for crop classification in general and cannabis in particular.

Predicted Mapping of Seabed Sediments Based on MBES Backscatter and Bathymetric Data: A Case Study in Joseph Bonaparte Gulf, Australia, Using Random Forest Decision Tree

Predictive mapping of seabed sediments based on multibeam bathymetric (BM), and backscatter (BS) data is effective for mapping the spatial distribution of the substrate. A robust modeling technique, the random forest decision tree (RFDT), was used to predict the seabed sediments in an area of the Joseph Bonaparte Gulf, Northern Australia, using the multibeam data and seabed sediment samples collected simultaneously. The results showed that: (1) Using multibeam bathymetry data in addition to multibeam backscatter data improves the prediction performance of the RFDT. In comparison to only multibeam backscatter data, the prediction performance achieved a ~10% improvement in sediment properties; it achieved a ~44.45% improvement of overall accuracy in sediment types, and a ~0.55 improvement in Kappa. (2) The underlying relationships between sediment properties and multibeam data show that there is an opposite non-linear correlation between sediment property-BS and sediment property-BM. For example, there is an obvious negative relationship between %mud-BS at incidence angles of 13° and 21°, but the relationship between %mud-BM is positive. As such, the RFDT is a useful and well-performing method in predicting the relationship between sediment properties and multibeam data and in predicting the distribution of sediment properties and types. However, the sediment prediction method in deep-water areas with high gravel content needs to be further evaluated.

Fusion of Sentinel-1 and Sentinel-2 data in mapping the impervious surfaces at city scale.

Urbanization creates new development in open spaces and agricultural fields, synonymous with increasing impervious surfaces. Such surfaces restrain the natural infiltration of water, and directly affect the non-point source pollution. Thus, consequential events like flooding and surface water degradation require spatial and quantitative information on impervious surfaces. Remote sensing technologies are widely used in impervious surface mapping of various geographical locations for environmental monitoring. In this study, the datasets from recently launched European Space Agency satellites (Sentinel-1 and Sentinel-2) and random forest classifier are used. The impervious surface growth of the study area, Lahore city, in 2015 and 2021, and growth trends are assessed. Results are validated with classification accuracy and comparison with similar datasets. The objective is to develop a reliable impervious surface mapping method with land cover quantification technique from multisource datasets. With a chi-square value of greater than 3.84 obtained from the McNemar test, the performance of fused data was superior to that of optical alone data in the classification. Over a 5-year period, Lahore grew at an annual rate of 2.14% comparable to the findings of Copernicus Land Services and the Atlas of Urban Expansion with an underestimation of 1% and 8.75%, respectively. Improvements in overall accuracy (2.7%) and kappa coefficient (5%) were seen in classified maps from fused datasets. Fusion of Sentinel datasets provide a reliable means of impervious surface mapping at city scale as an indicator of environmental quality which is valuable for the sustainable management of the city.

Predicting the clinical management of skin lesions using deep learning

Automated machine learning approaches to skin lesion diagnosis from images are approaching dermatologist-level performance. However, current machine learning approaches that suggest management decisions rely on predicting the underlying skin condition to infer a management decision without considering the variability of management decisions that may exist within a single condition. We present the first work to explore image-based prediction of clinical management decisions directly without explicitly predicting the diagnosis. In particular, we use clinical and dermoscopic images of skin lesions along with patient metadata from the Interactive Atlas of Dermoscopy dataset (1011 cases; 20 disease labels; 3 management decisions) and demonstrate that predicting management labels directly is more accurate than predicting the diagnosis and then inferring the management decision (13.73 pm 3.93% and 6.59 pm 2.86% improvement in overall accuracy and AUROC respectively), statistically significant at p < 0.001. Directly predicting management decisions also considerably reduces the over-excision rate as compared to management decisions inferred from diagnosis predictions (24.56% fewer cases wrongly predicted to be excised). Furthermore, we show that training a model to also simultaneously predict the seven-point criteria and the diagnosis of skin lesions yields an even higher accuracy (improvements of 4.68 pm 1.89% and 2.24 pm 2.04% in overall accuracy and AUROC respectively) of management predictions. Finally, we demonstrate our model’s generalizability by evaluating on the publicly available MClass-D dataset and show that our model agrees with the clinical management recommendations of 157 dermatologists as much as they agree amongst each other.

Plant species classification in salt marshes using phenological parameters derived from Sentinel-2 pixel-differential time-series

Salt marshes are one of the most productive but vulnerable ecosystems on Earth; and due to the continued intensification of natural and anthropogenic pressures on them, accurate and timely information on the distribution of plant species in salt marshes is needed for effective coastal management. Time-series approaches have been widely applied to classify plant species; however, developing time-series with high spatial and temporal resolution over coastal zones remains challenging due to the influence caused by frequent cloud cover and periodic tidal fluctuations. In this study, aiming at the above challenges, we presented a saltmarsh vegetation classification method using phenological parameters derived from Sentinel-2 pixel-differential time-series (PDTS): first, a PDTS that each pixel has a different distribution of observations was constructed using a time-series cloud mask; second, a tidal filter determined by the threshold and frequency of the modified normalized difference water index (MNDWI) was used to exclude tide-related observations from the PDTS; third, phenological parameters that highlight the differences among salt marsh vegetation were extracted from a two-term Fourier fitting curve as classification features; and finally, the random forest algorithm was used for plant species classification with the assistance of sample data. Six common plant species (Spartina alterniflora, Phragmites australis, Suaeda salsa, Tamarix chinensis, Imperata cylindrica, and Scirpus mariqueter) from three representative coastal sites in China were analyzed. The major results were as follows: (1) The MNDWI demonstrated superior ability in identifying flooding pixels, with an overall accuracy of ~0.91. After tidal filtering, the R2 of the fitting curve for more than 70% of the vegetated salt marsh pixels was improved with an average increase of 0.113. (2) The six plant species exhibit unique phenological characteristics. In particular, P. australis has an advanced green-up season, 19–42 days earlier than that of the other plant species mentioned above, whereas S. alterniflora senesces one to two months later than the native plant species. (3) The average overall accuracy of the plant species classification based on the PDTS was 81.5%. Compared with a single-image-based classification, the PDTS-based classification demonstrated a ~ 5.1% improvement in overall accuracy, which is expected to serve the annual monitoring dynamics of the salt marsh.