General Workflow Research Articles

Scalable modular synthetic data generation for advancing aerial autonomy

One major barrier to advancing aerial autonomy has been collecting large-scale aerial datasets for training machine learning models. Due to costly and time-consuming real-world data collection through deploying drones, there has been an increasing shift towards using synthetic data for training models in drone applications. However, to increase widespread generalization and transferring models to real-world, increasing the diversity of simulation environments to train a model over all the varieties and augmenting the training data, has been proved to be essential. Current synthetic aerial data generation tools either lack data augmentation or rely heavily on manual workload or real samples for configuring and generating diverse realistic simulation scenes for data collection. These dependencies limit scalability of the data generation workflow. Accordingly, there is a major challenge in balancing generalizability and scalability in synthetic data generation. To address these gaps, we introduce a scalable Aerial Synthetic Data Augmentation (ASDA) framework tailored to aerial autonomy applications. ASDA extends a central data collection engine with two scriptable pipelines that automatically perform scene and data augmentations to generate diverse aerial datasets for different training tasks. ASDA improves data generation workflow efficiency by providing a unified prompt-based interface over integrated pipelines for flexible control. The procedural generative approach of our data augmentation is performant and adaptable to different simulation environments, training tasks and data collection needs. We demonstrate the effectiveness of our method in automatically generating diverse datasets and show its potential for downstream performance optimization. Our work contributes to generating enhanced benchmark datasets for training models that can generalize better to real-world situations.Video:youtube.com/watch?v=eKpOh-K-NfQ

Towards an optimized design for an elevated cyclonic home using numerical simulations and active learning framework

This research presents a general optimization workflow and a geometrical parametric study using computational fluid dynamics (CFD) analyses in conjunction with a Kriging-based active learning strategy. Motivated by the concept of a wind-resistant “cyclonic home”, the present work conducts a systematic numerical wind study for a typical gable-roof elevated house structure across a parameter space consisting of varying incident wind directions and roof pitch angles. The house model is fully parametrized such that geometrical and mesh generation are automatically performed to facilitate the deployment of active learning for establishing a surrogate model of wind-induced forces. A wind force map is produced which reveals the relationship between the incident wind angle and roof pitch. From this, statistical analysis exposes which roof pitch angles offer greater wind-resistance, i.e., a roof pitch of θ=32.9° is found to provide the lowest mean force across all possible wind angles, whereas roof pitches of θ=12.7° and θ=45.0° yield the largest forces. The influence of the overhanging eave, a typical feature of such houses, is varied to understand its role on overall roof performance. Quartering winds are found to provide a greater contribution to overall roof forces as the eave length is extended. In contrast, for winds perpendicular to the roof ridge, its extension provides a relatively stable total force. These findings can assist a variety of stakeholders from engineering and risk assessment disciplines in pursuit of buildings resilient to high-winds due to tropical cyclones.

A rapid procedure to generate stably transfected HEK293 suspension cells for recombinant protein manufacturing: Yield improvements, bioreactor production and downstream processing

The human cell line HEK293 is one of the preferred choices for manufacturing therapeutic proteins and viral vectors for human applications. Despite its increased use, it is still considered in disadvantage in production aspects compared to cell lines such as the CHO cell line. We provide here a simple workflow for the rapid generation of stably transfected HEK293 cells expressing an engineered variant of the SARS-CoV-2 Receptor Binding Domain (RBD) carrying a coupling domain for linkage to VLPs through a bacterial transpeptidase-sortase (SrtA). To generate stable suspension cells expressing the RBD-SrtA, a single two plasmids transfection was performed, with hygromycin selection. The suspension HEK293 were grown in adherent conditions, with 20% FBS supplementation. These transfection conditions increased cell survival, allowing the selection of stable cell pools, which was otherwise not possible with standard procedures in suspension. Six pools were isolated, expanded and successfully re-adapted to suspension with a gradual increase of serum-free media and agitation. The complete process lasted four weeks. Stable expression with viability over 98% was verified for over two months in culture, with cell passages every 4–5 days. With process intensification, RBD-SrtA yields reached 6.4 μg/mL and 13.4 μg/mL in fed-batch and perfusion-like cultures, respectively. RBD-SrtA was further produced in fed-batch stirred tank 1L-bioreactors, reaching 10-fold higher yields than perfusion flasks. The trimeric antigen displayed the conformational structure and functionality expected. This work provides a series of steps for stable cell pool development using suspension HEK293 cells aimed at the scalable production of recombinant proteins.

Bioorthogonal Chemical Imaging of Cell Metabolism Regulated by Aromatic Amino Acids.

Essential aromatic amino acids (AAAs) are building blocks for synthesizing new biomasses in cells and sustaining normal biological functions. For example, an abundant supply of AAAs is important for cancer cells to maintain their rapid growth and division. With this, there is a rising demand for a highly specific, noninvasive imaging approach with minimal sample preparation to directly visualize how cells harness AAAs for their metabolism in situ. Here, we develop an optical imaging platform that combines deuterium oxide (D2O) probing with stimulated Raman scattering (DO-SRS) and integrates DO-SRS with two-photon excitation fluorescence (2PEF) into a single microscope to directly visualize the metabolic activities of HeLa cells under AAA regulation. Collectively, the DO-SRS platform provides high spatial resolution and specificity of newly synthesized proteins and lipids in single HeLa cell units. In addition, the 2PEF modality can detect autofluorescence signals of nicotinamide adenine dinucleotide (NADH) and Flavin in a label-free manner. The imaging system described here is compatible with both in vitro and in vivo models, which is flexible for various experiments. The general workflow of this protocol includes cell culture, culture media preparation, cell synchronization, cell fixation, and sample imaging with DO-SRS and 2PEF modalities.

Generic Workflow to Predict Medicine Concentrations in Human Milk Using Physiologically-Based Pharmacokinetic (PBPK) Modelling-A Contribution from the ConcePTION Project.

Women commonly take medication during lactation. Currently, there is little information about the exposure-related safety of maternal medicines for breastfed infants. The aim was to explore the performance of a generic physiologically-based pharmacokinetic (PBPK) model to predict concentrations in human milk for ten physiochemically diverse medicines. First, PBPK models were developed for "non-lactating" adult individuals in PK-Sim/MoBi v9.1 (Open Systems Pharmacology). The PBPK models predicted the area-under-the-curve (AUC) and maximum concentrations (Cmax) in plasma within a two-fold error. Next, the PBPK models were extended to include lactation physiology. Plasma and human milk concentrations were simulated for a three-months postpartum population, and the corresponding AUC-based milk-to-plasma (M/P) ratios and relative infant doses were calculated. The lactation PBPK models resulted in reasonable predictions for eight medicines, while an overprediction of human milk concentrations and M/P ratios (>2-fold) was observed for two medicines. From a safety perspective, none of the models resulted in underpredictions of observed human milk concentrations. The present effort resulted in a generic workflow to predict medicine concentrations in human milk. This generic PBPK model represents an important step towards an evidence-based safety assessment of maternal medication during lactation, applicable in an early drug development stage.

Automated generation of scientific workflow generators with WfChef

Scientific workflow applications have gained significant importance, and their automated and efficient execution on large-scale computing platforms has been the subject of extensive research and development. For these efforts to be successful, a solid experimental methodology is needed to evaluate workflow algorithms and systems. A foundation for this methodology is the availability of realistic workflow instances. Although public repositories provide workflow instances for a few scientific applications, these are limited in scope, and workflow instances are not available for all application scales of interest. To address this limitation, previous work has developed generators of synthetic workflow instances of arbitrary scales. Despite being popular, the implementation of these generators is a manual and labor-intensive process that requires expert application knowledge. As a result, these generators only target a handful of applications, even though there are hundreds of workflow applications in production.We introduce WfChef , a fully automated framework for constructing a synthetic workflow generator for any scientific application. Based on an input set of workflow instances for a particular application, WfChef automatically produces a synthetic workflow generator. To measure the realism of the generated workflows, we define and evaluate several metrics. Using these metrics, we compare the realism of the workflows generated by WfChef generators to that of the workflows generated by the previously available, hand-crafted generators. We find that WfChef generators not only require zero development effort (because they are automatically produced), but also generate workflows that are more realistic than those generated by hand-crafted generators.

Application of population physiologically based pharmacokinetic modelling to optimize target expression and clearance mechanisms of therapeutic monoclonal antibodies.

To use population physiologically based pharmacokinetic (PopPBPK) modelling to optimize target expression, kinetics and clearance of HER1/2 directed therapeutic monoclonal antibodies (mAbs). Thus, to propose a general workflow of PopPBPK modelling and its application in clinical pharmacology. Full PBPK model of pertuzumab (PTZ) was developed in patient population using Simcyp V21R1 incorporating mechanistic targeted-mediated drug disposition process by fitting known clinical PK and sparse receptor proteomics data to optimize target expression and kinetics of HER2 receptor. Trastuzumab (TTZ) PBPK modelling was used to validate the optimized HER2 target. Additionally, the simulator was also used to develop a full PBPK model for the HER1-directed mAb cetuximab (CTX) to assess the underlying targeted-mediated drug disposition-independent elimination mechanisms. HER2 final parameterisation coming from the PBPK modelling of PTZ was successfully cross validated through PBPK modelling of TTZ with average fold error (AFE), absolute AFE and percent prediction error values for area under the concentration-time curve (AUC) and maximum plasma concentration (Cmax ) of 1.13, 1.16 and 16, and 1.01, 1.07 and 7, respectively. CTX PBPK model performance was validated after the incorporation of an additional systemic clearance of 0.033 L/h as AFE and absolute AFE showed an acceptable predictive power of AUC and Cmax with percent prediction error of 13% for AUC and 10% for Cmax . Optimisation of both system and drug related parameters were performed through PBPK modelling to improve model performance of therapeutic mAbs (PTZ, TTZ and CTX). General workflow was proposed to develop and apply PopPBPK to support clinical development of mAbs targeting same receptor.

Genomes, Transcriptomes, Proteomes, Metabolomes in Bioinformatics Investigating Alzheimers Disease

Alzheimers disease is affecting many people, especially people older than 65, living in the world. However, the cause and cure of these disease continue to confuse the scientists. Omics analysis, which includes four levels: genomes, transcriptomes, proteomes and metabolomes analysis, can dig out the information lying behind large amount of data and its really useful for complex disease like AD (Alzheimers Disease). In this article, I talked about the four levels of omics analysis investigating AD in details. In each level, the general workflow, the information we can get and the examples in AD have been involved. This paper will give the readers a general view of the function of omics analysis in AD.

The pysat ecosystem

The Python Satellite Data Analysis Toolkit (pysat) is an open source package that implements a general data analysis workflow for arbitrary data sets, providing a consistent manner for obtaining, managing, analysing, and processing data, including modelled and observational ground and space-based data sets for the space sciences. Pysat enables systematic and individual treatment of data as well as simplifies rigorous data access and use, allowing larger-scale scientific efforts including machine learning, data assimilation, and constellation instrumentation processing. Since the start of its development pysat has evolved into an ecosystem, separating general file and data handling functionality from both individual data set support and generalized data analysis. This design choice ensures that the core pysat package has only the necessary functionality required to provide data management services for the wider development community. The shift of data and analysis support to ecosystem packages makes it easier for the community to contribute to, as well as use, the full array of features and data sources enabled by pysat. Pysat’s ease of use, and generality, supports adoption outside of professional science to include industry, citizen science, and education.

Establishing structure–property linkages for wicking time predictions in porous polymeric membranes using a data-driven approach

The goal of this study is to develop correlations between microstructure morphology and macroscopic material behavior, known as structure–property linkages. These correlations can be used to predict material behavior and enable virtual materials design efforts. In this work the structure–property linkages for the capillary-driven fluid transport through highly porous open-pored polymeric membranes are determined by a data-driven approach. To establish linkages, about 400 porous microstructures with different geometrical features are algorithmically generated and characterized in 3D, using fluid flow simulations and image analysis methods. The data processing pipeline for the generation and analysis of the microstructures is implemented by a generic workflow tool called KadiStudio, which is embedded in the research data infrastructure Kadi4mat. The data-driven analysis enables predictions about the propagation time of a fluid over definable distances when only the porosity and the ligament radius are known as microstructural properties. The generated knowledge can be utilized for an accelerated development of novel polymeric membranes with an optimized pore structure.

Multiplatform untargeted metabolomics.

Metabolomics samples like human urine or serum contain upwards of a few thousand metabolites, but individual analytical techniques can only characterize a few hundred metabolites at best. The uncertainty in metabolite identification commonly encountered in untargeted metabolomics adds to this low coverage problem. A multiplatform (multiple analytical techniques) approach can improve upon the number of metabolites reliably detected and correctly assigned. This can be further improved by applying synergistic sample preparation along with the use of combinatorial or sequential non-destructive and destructive techniques. Similarly, peak detection and metabolite identification strategies that employ multiple probabilistic approaches have led to better annotation decisions. Applying these techniques also addresses the issues of reproducibility found in single platform methods. Nevertheless, the analysis of large data sets from disparate analytical techniques presents unique challenges. While the general data processing workflow is similar across multiple platforms, many software packages are only fully capable of processing data types from a single analytical instrument. Traditional statistical methods such as principal component analysis were not designed to handle multiple, distinct data sets. Instead, multivariate analysis requires multiblock or other model types for understanding the contribution from multiple instruments. This review summarizes the advantages, limitations, and recent achievements of a multiplatform approach to untargeted metabolomics.

Wisdom of Two Crowds: Misinformation Moderation on Reddit and How to Improve this Process---A Case Study of COVID-19

Past work has explored various ways for online platforms to leverage crowd wisdom for misinformation detection and moderation. Yet, platforms often relegate governance to their communities, and limited research has been done from the perspective of these communities and their moderators. How is misinformation currently moderated in online communities that are heavily self-governed? What role does the crowd play in this process, and how can this process be improved? In this study, we answer these questions through semi-structured interviews with Reddit moderators. We focus on a case study of COVID-19 misinformation. First, our analysis identifies a general moderation workflow model encompassing various processes participants use for handling COVID-19 misinformation. Further, we show that the moderation workflow revolves around three elements: content facticity, user intent, and perceived harm. Next, our interviews reveal that Reddit moderators rely on two types of crowd wisdom for misinformation detection. Almost all participants are heavily reliant on reports from crowds of ordinary users to identify potential misinformation. A second crowd--participants' own moderation teams and expert moderators of other communities--provide support when participants encounter difficult, ambiguous cases. Finally, we use design probes to better understand how different types of crowd signals---from ordinary users and moderators---readily available on Reddit can assist moderators with identifying misinformation. We observe that nearly half of all participants preferred these cues over labels from expert fact-checkers because these cues can help them discern user intent. Additionally, a quarter of the participants distrust professional fact-checkers, raising important concerns about misinformation moderation.

An Optimized Workflow for Digital Surface Model Series Generation Based on Historical Aerial Images: Testing and Quality Assessment in the Beach-Dune System of Sa Ràpita-Es Trenc (Mallorca, Spain)

We propose an optimized Structure-from-Motion (SfM) Multi-View Stereopsis (MVS) workflow, based on minimizing different errors and inaccuracies of historical aerial photograph series (1945, 1979, 1984, and 2008 surveys), prior to generation of elevation-calibrated historical Digital Surface Models (hDSM) at 1 m resolution. We applied LiDAR techniques on Airborne Laser Scanning (ALS) point clouds (Spanish PNOA LiDAR flights of 2014 and 2019) for comparison and validation purposes. Implementation of these products in multi-temporal analysis requires quality control due to the diversity of sources and technologies involved. To accomplish this, (i) we used the Mean Absolute Error (MAE) between GNSS-Validation Points and the elevations observed by DSM-ALS to evaluate the elevation accuracy of DSM-ALS generated with the LAScatalog processing engine; (ii) optimization of the SfM sparse clouds in the georeferencing step was evaluated by calculating the Root Mean Square Error (RMSE) between the Check Points extracted from DSM-ALS and the predicted elevations per sparse cloud; (iii) the MVS clouds were evaluated by calculating the MAE between ALS-Validation Points and the predicted elevations per MVS cloud; iv) the accuracy of the resulting historical SfM-MVS DSMs were assessed using the MAE between ALS-Validation Points and the observed elevations per historical DSM; and (v) we implemented a calibration method based on a linear correction to reduce the elevation discrepancies between historical DSMs and the DSM-ALS 2019 reference elevations. This optimized workflow can generate high-resolution (1 m pixel size) hDSMs with reasonable accuracy: MAE in z ranges from 0.41 m (2008 DSM) to 5.21 m (1945 DSM). Overall, hDSMs generated using historical images have great potential for geo-environmental processes monitoring in different ecosystems and, in some cases (i.e., sufficient image overlapping and quality), being an acceptable replacement for LiDAR data when it is not available.

Industry Perspective on the Pharmacokinetic and ADME Characterization of Heterobifunctional Protein Degraders.

Targeted protein degraders (TPDs), specifically the bifunctional protein degraders discussed in this manuscript, consist of two linked ligands for a protein of interest and an E3 ligase, resulting in molecules which largely violate accepted physicochemical limits (e.g. Lipinski's Rule of 5) for oral bioavailability. In 2021, the IQ Consortium Degrader DMPK/ADME Working Group undertook a survey of eighteen IQ member and non-member companies working on degraders to understand whether the characterization and optimization of these molecules were different from any other beyond the Rule of 5 (bRo5) compounds. Additionally, the working group sought to identify PK/ADME areas in need of further evaluation and where additional tools could aid in more rapid advancement of TPDs to patients. The survey revealed that although TPDs reside in a challenging bRo5 physicochemical space, most respondents focus their efforts on oral delivery. Physicochemical properties required for oral bioavailability were generally consistent across the companies surveyed. Many of the member companies used modified assays to address challenging degrader properties (e.g. solubility, non-specific binding), but only half indicated that they modified their drug discovery workflows. The survey also suggested the need for further scientific investigation in the areas of CNS penetration, active transport, renal elimination, lymphatic absorption, in silico/machine learning, and human pharmacokinetic prediction. Based on the survey results, the Degrader DMPK/ADME Working Group concludes that TPD evaluation does not fundamentally differ from other bRo5 compounds, but requires some modification compared to traditional small molecules and proposes a generic workflow for PK/ADME evaluation of bifunctional TPDs. Significance Statement Based on an industry survey, this article provides an understanding of the current state of ADME science pertaining to characterizing and optimizing targeted protein degraders, specifically bifunctional protein degraders, based upon the responses provided by 18 IQ consortium members and non-members developing targeted protein degraders to treat disease. Additionally, this article puts into context the differences / similarities in methods and strategies utilized for heterobifunctional protein degraders compared to other bRo5 molecules and those used for conventional small molecule drugs.

A machine learning approach for digital watermarking

ABSTRACT Recently, machine learning (ML) has been applying in almost all scientific fields to model and simulate the behaviour of complex systems. At the same time, the number of the proposed watermarking techniques have been increasing every year. Although most of the researchers in the watermarking and data hiding field put all their effort to compete each other to develop more efficient algorithm, the selection of the most efficient one is almost impossible for industries or software developers. In other words, comparison among all the watermarking techniques in order to select one watermarking technique would be cumbersome task. Moreover, the computer security is becoming more advanced which a single watermark technique cannot fulfil all the required criteria and there is a chance that an anti-watermarking is developed to remove the embedded watermark from the host data. This gap of knowledge to use the watermarking technology for a highly secure application is a real nightmare for the information security engineers and software designers. In this paper, two new approaches are proposed to train deep learning and shallow learning models based on the state-of-the-arts watermarking techniques. For these proposed approaches, several ML algorithms are applied to model various watermarked data, which are watermarked by different watermarking techniques in various spectrums and domains. An experimental setup is constructed based on several speeches, audios, images and videos beside the Amazon Sagemaker as ML modelling to implement the proposed approach. The experimental results show that apart from an overall effectiveness of the model, it would resolve some ambiguities by applying the proposed ML approach as a general watermarking workflow.

General Practices’ Experiences With Patients’ Web-Based Access to Medical Records: Survey Study

Patients' web-based access to their medical records is expected to promote their role and responsibility in managing their own health and treatments and supporting shared decision-making. As of July 2020, general practices in the Netherlands are legally obliged to provide their patients access to their electronic medical records. Web-based access provision is facilitated and stimulated through a national support program named OPEN. We aimed to investigate general practice staff experiences with providing web-based access; investigate its impact on patient consultations, administrative actions, and patient inquiries; and investigate how it affects routine general practice workflow processes. In October 2021, a total of 3813 general practices in the Netherlands were invited to complete a web-based survey that included questions regarding their experiences with the provision of web-based access to medical records and how it affects routine general practice workflow. Responses of general practices that started providing web-based access before 2020, in 2020, or in 2021 were analyzed to identify trends. Of 3813 invited general practices, 523 (13.72%) completed the survey. Approximately all responding general practices (487/523, 93.1%) indicated that they provide web-based access. Experiences with patients' web-based access were diverse, with 36.9% (178/482) primarily positive, 8.1% (39/482) primarily negative, 42.3% (204/482) neutral, and 12.7% (61/482) could not (yet) indicate how they experienced web-based access. Of the total, two-thirds (311/473, 65.8%) reported an increase in e-consultations and a similar percentage (302/474, 63.7%) indicated an increase in administrative actions associated with web-based access provision. A small proportion of the practices (≤10%) experienced a decrease in patient contacts. Earlier adoption of web-based access was associated with a more positive attitude toward web-based access and more positive experienced effects related to patient contacts and general practice workflow. The surveyed general practices mainly experienced providing web-based access as either neutral or mostly positive, despite an increased number of patient contacts and administrative burden that were associated with its adoption. Periodic monitoring of experiences is needed to understand the temporal or structural nature of both the intended and unintended effects of patients' web-based access to medical records for general practices and their staff.

Computer Vision Applications In Construction And Asset Management Phases: A Literature Review

Recent advances in digital photography and unmanned aerial vehicle (UAV) platforms make visual data from construction project sites more accessible to project teams. To semi-automatically or automatically obtain the essential information, evaluate the ongoing activities or operations, and address project-level challenges, researchers have focused on applying various computer vision (CV)-based methods to process and interpret the acquired visual data. This research developed a framework to summarize the vision-based methods that have been applied to construction/asset management operations through a systematic literature review. The reviewed literature was composed of 103 journal papers from 2011 to 2020. All the reviewed journal papers were from the Ei Compendex database with specific search criteria. The developed framework consisted of two parts: use cases and CV domains. Use cases contained five aspects: safety monitoring, productivity improvement, progress monitoring, infrastructure inspection, and robotic application. CV domains contained six aspects: image processing, object classification, object detection, object tracking, pose estimation, and 3D reconstruction. All eleven aspects were integrated from the reviewed papers. For each reviewed paper, the general workflow of applied vision-based approaches was described and categorized into each use case. A trending timeline was developed to analyze the popularity of the identified use cases and CV domains within the reviewed time period. Both the quantity and variety of construction use cases and CV domains have increased. Challenges and limitations of applying CV-based methods in the construction industry were also identified. This paper provides readers with a summary of how CV-based methods have been used in the construction industry and serve as a reference for future research and development.

3D least-squares reverse time migration in VTI media based on pseudoacoustic wave equation and multi-GPU parallel acceleration

As the high-density seismic acquisition is progressively becoming an essential mode of seismic exploration, developing high-precision imaging methods for 3D datasets has become an urgent industry requirement. Unfortunately, the two-dimensional least-squares reverse time migration (LSRTM) can hardly deal with the influence of the third dimension because the structure of reflectors is constantly changing in nearly all directions. Furthermore, anisotropy is very common in underground media, which makes the LSRTM based on isotropic assumption fail to locate the reflectors accurately and seriously limits the identification of geological stratification. Considering the above two problems, this paper derives the least-squares reverse time migration in three-dimensional vertical transversely isotropic media (3D-VTI-LSRTM) and develops a workflow which can realize 3D-LSRTM in VTI media based on this theory. Our workflow uses multi-GPU acceleration and heterogeneous CPU/GPU parallelism. Compared with the general parallel CPU workflow or single GPU workflow, it is feasible that the GPU application significantly improves computing efficiency in practical applications. In addition, this paper extends VTI-LSRTM from 2D to 3D, which solves the problem that the 2D algorithm cannot eliminate the lateral direction reflection artifacts. For the self-made 3D-VTI depression model and the 3D-VTI-SEG/EAGE salt model, we use four NVIDIA Tesla k40c graphics cards for testing. The results of the tests are correct and significantly improved compared to the 3D-VTI-RTM method.

Improving Deep Learning Classifiers Performance via Preprocessing and Class Imbalance Approaches in a Plant Disease Detection Pipeline

The foundation of effectively predicting plant disease in the early stage using deep learning algorithms is ideal for addressing food insecurity, inevitably drawing researchers and agricultural specialists to contribute to its effectiveness. The input preprocessor, abnormalities of the data (i.e., incomplete and nonexistent features, class imbalance), classifier, and decision explanation are typical components of a plant disease detection pipeline based on deep learning that accepts an image as input and outputs a diagnosis. Data sets related to plant diseases frequently display a magnitude imbalance due to the scarcity of disease outbreaks in real field conditions. This study examines the effects of several preprocessing methods and class imbalance approaches and deep learning classifiers steps in the pipeline for detecting plant diseases on our data set. We notably want to evaluate if additional preprocessing and effective handling of data inconsistencies in the plant disease pipeline may considerably assist deep learning classifiers. The evaluation’s findings indicate that contrast limited adaptive histogram equalization (CLAHE) combined with image sharpening and generative adversarial networks (GANs)-based approach for resampling performed the best among the preprocessing and resampling techniques, with an average classification accuracy of 97.69% and an average F1-score of 97.62% when fed through a ResNet-50 as the deep learning classifier. Lastly, this study provides a general workflow of a disease detection system that allows each component to be individually focused on depending on necessity.

Hard Potato: A Python Library to Control Commercial Potentiostats and to Automate Electrochemical Experiments.

Here, we develop and show the use of an open-source Python library to control commercial potentiostats. It standardizes the commands for different potentiostat models, opening the possibility to perform automated experiments independently of the instrument used. At the time of this writing, we have included potentiostats from CH Instruments (models 1205B, 1242B, 601E, and 760E) and PalmSens (model Emstat Pico), although the open-source nature of the library allows for more to be included in the future. To showcase the general workflow and implementation of a real experiment, we have automated the Randles-Ševčı́k methodology to determine the diffusion coefficient of a redox-active species in solution using cyclic voltammetry. This was accomplished by writing a Python script that includes data acquisition, data analysis, and simulation. The total run time was 1 min and 40 s, well below the time it would take even an experienced electrochemist to apply the methodology in a traditional manner. Our library has potential applications that expand beyond the automation of simple repetitive tasks; for example, it can interface with peripheral hardware and well-established third-party Python libraries as part of a more complex and intelligent setup that relies on laboratory automation, advanced optimization, and machine learning.