Hybrid Dataset Research Articles

Improved Speech Authenticity Detection in Chinese-English Bilingual Contexts.

The rapid evolution of voice technology has heightened the need for robust detection systems to distinguish between authentic and tampered speech. Recent competitions have significantly advanced the development of countermeasures against spoofing attacks. However, while advancements in detection technologies have been notable, existing methods often focus on a single type of tampering and language. Our contribution lies in developing an improved model that integrates an enhanced ResNet architecture with an LSTM to improve the detection of tampered audio, particularly in challenging multilingual scenarios. In the experiments, we built a hybrid dataset from self-recording Chinese speech and public VCTK2 English samples, enhanced the ResNet model generalization capabilities, and evaluated our approach using the bilingual dataset. Experiment results demonstrate that the proposed approach achieves a superior performance with an equal error rate of 11.62%, even in the face of bilingual conditions, and, more importantly, outperforms the leading models from ASVSpoof 2021 and ADD 2022 competitions. We also employed advanced tampering techniques, including CycleGAN voice conversion and auto splicing, to simulate real-world tampering scenarios and verify the effectiveness of the proposed approach.

An alloy-agnostic machine learning framework for process mapping in laser powder bed fusion

PurposeThis study aims to introduce an image-based method to determine the processing window for a given alloy system using laser powder bed fusion equipment based on achieving the desired melting mode across multiple materials for powder-free specimens. The method uses a convolutional neural network trained to classify different track morphologies across different alloy systems to select appropriate printing settings. This method is intended for the development of new alloy systems, where the powder feedstock may be unavailable, or prohibitively expensive to manufacture.Design/methodology/approachA convolutional neural network is designed from scratch to identify the 4 key melting modes that are observed in laser powder bed fusion additive manufacturing across different alloy systems. To increase the prediction accuracy and generalisation accuracy across different materials, the network is trained using a novel hybrid data set that combines fully unsupervised learning with semi-supervised learning.FindingsThis study demonstrates that our convolutional network with a novel hybrid training approach can be generalised across different materials, and k-fold validation shows that the model retains good accuracy with changing training conditions. The model can predict the processing maps for the different alloys with an accuracy of up to 96% in some cases. It is also shown that powder-free single-track experiments are a useful indicator for predicting the final print quality of a component.Originality/valueThe “invariant information clustering” (IIC) approach is applied to process optimisation for additive manufacturing, and a novel hybrid data set construction approach that accounts for uncertainty in the ground truth data, enables the trained convolutional model to perform across a range of different materials and most importantly, generalise to materials outside of the training data set. Compared to the traditional cross-sectioning approach, this method considers the whole length of the single track when determining the melting mode.

Evaluating the Impact of BoNT-A Injections on Facial Expressions: A Deep Learning Analysis.

Botulinum Toxin Type A (BoNT-A) injections are widely used for facial rejuvenation, but their effects on facial expressions remain unclear. This study aims to objectively measure the impact of BoNT-A injections on facial expressions using deep learning techniques. 180 patients aged 25-60 years who underwent BoNT-A application to the upper face were included. Patients were photographed with neutral, happy, surprised, and angry expressions before and 14 days after the procedure. A Convolutional Neural Network (CNN)-based Facial Emotion Recognition (FER) system analyzed 1440 photographs using a hybrid dataset of clinical images and the Karolinska Directed Emotional Faces (KDEF) dataset. The CNN model accurately predicted 90.15% of the test images. Significant decreases in the recognition of angry and surprised expressions were observed post-injection (p<0.05), with no significant changes in happy and neutral expressions (p>0.05). Angry expressions were often misclassified as neutral or happy (p<0.05), and surprised expressions were more likely to be perceived as neutral (p<0.05). Deep learning can effectively assess the impact of BoNT-A injections on facial expressions, providing more standardized data than traditional surveys. BoNT-A may reduce the expression of anger and surprise, potentially leading to a more positive facial appearance and emotional state. Further studies are needed to understand the broader implications of these changes.

Chemical-guided screening of top-performing metal–organic frameworks for hydrogen storage: An explainable deep attention convolutional model

Metal-organic framework (MOF)-based adsorptive hydrogen storage holds promise for enhancing the sustainable design of hydrogen storages by enhancing the usable volumetric (UV) and gravimetric (UG) capacities. However, the extensive number of MOFs poses a challenge in the search for optimal materials owing to the lack of an efficient and interpretable high-throughput screening method. This study introduces an explainable artificial intelligence (XAI) framework to expedite the discovery of high-capacity hydrogen adsorbents by predicting the UV and UG capacities using an attention densely connected convolutional (ADCC) network. A new hybrid dataset with various operating conditions and comprising 24 physical–chemical descriptors, such as void fraction (VF) and metal percentage (MP), was utilized to develop the ADCC model. The explainable ADCC model demonstrated superior predictive performance for the UV and UG capacities, with R2 values of 0.9886 and 0.9982, respectively. The inclusion of the chemical descriptors MOFs enhanced the prediction accuracy of the ADCC model. The XAI analysis showed that VF and MP dominated physical and chemical descriptors, respectively, for UV and UG. Consequently, the ADCC model identified EFAYIU—a real MOF—as a promising hydrogen storage material with UV and UG capacities of 51.55 g H2/L and 11.37 wt%, respectively, surpassing the current materials for hydrogen storage. Additionally, the identified EFAYIU was validated based on molecular simulations, confirming the high hydrogen adsorptive capacities obtained by the ADCC model. Thus, the proposed AI-based high-throughput screening method enables the rapid discovery of high-performance MOFs for sustainable hydrogen storage.

Development of a machine learning model to improve estimates of material stock and embodied emissions of roads

Material flow analysis is an important tool for estimating material flows and embedded emissions of transport infrastructure. Missing attributes tend to be a major barrier to accurate estimates. In this study a machine learning model is developed to estimate the missing data in a statistics dataset of roads, to enable a bottom-up material stock and flow analysis. The proposed approach was applied to the Swedish road network to predict missing data for road width in the statistical dataset. The predicted hybrid dataset was then used to estimate material stocks, flows, and embodied emissions from Year 2020 to Year 2045 using decarbonization scenarios with a supply chain perspective. The study demonstrates that machine learning models can be used to enable national-level material stock and flow analyses of roads. Multiple machine learning algorithms were tested, and the best performing model achieved an R2 value of 0.784. In the scenario-based analysis, the embodied emissions of Swedish roads could be reduced by up to 51% using available materials.

Predicting hurricane evacuation behavior synthesizing data from travel surveys and social media

Evacuation behavior models estimated using post-disaster surveys are not adequate to predict real-time dynamic population response as a hurricane unfolds. With the emergence of ubiquitous technology and devices in recent times, social media data with its higher spatio-temporal coverage has become a potential alternative for understanding evacuation behaviour during hurricanes. However, these data are often associated with selection bias and population representativeness issues. To that extent, the current study proposes a novel data fusion algorithm to combine heterogeneous data sources from transportation systems and social media, in a unified framework to understand and predict real-time population response during hurricanes. Specifically, Twitter data of 2300 users are collected for evacuation response during Hurricane Irma and augmented behaviourally (probabilistically) with a representative National Household Travel Survey (NHTS) data, thus creating a hybrid dataset to improve the representativeness as well as provide a rich set of explanatory variables for understanding the evacuation behavior. The fusion process is conducted using a probabilistic matching method based on a set of common attributes across NHTS and Twitter. The fused dataset is employed to estimate the evacuation model and a comparison exercise is conducted to evaluate the performance of the model via fusion. The model fitness measures clearly demonstrate the improvement in data fit for the evacuation model through the proposed fusion algorithm. Further, we conduct a prediction assessment to illustrate the applicability of the proposed fusion technique and the results clearly highlight the improvement in the evacuation prediction rate achieved through the fused models. The proposed data-driven methods will enhance our ability to predict time-dependent evacuation demand for better hurricane response operations such as targeted warning dissemination and improved evacuation traffic management, allowing emergency plans to be more adaptive.

Hybrid Data Augmentation for Enhanced Crack Detection in Building Construction

Quality management in construction projects necessitates early defect detection, traditionally conducted manually by supervisors, resulting in inefficiencies and human errors. Addressing this challenge, research has delved into automating defect detection using computer vision technology, yet progress has been impeded by data limitations. Numerous studies have explored generating virtual images to tackle this issue. However, these endeavors have fallen short in providing image data adaptable to detecting defects amidst evolving on-site construction conditions. This study aims to surmount this obstacle by constructing a hybrid dataset that amalgamates virtual image data with real-world data, thereby enhancing the accuracy of deep learning models. Virtual images and mask images for the model are concurrently generated through a 3D virtual environment and automatic rendering algorithm. Virtual image data are built by employing a developed annotation system to automatically annotate through mask images. This method improved efficiency by automating the process from virtual image creation to annotation. Furthermore, this research has employed a hierarchical classification system in generating virtual image datasets to reflect the different types of defects that can occur. Experimental findings demonstrate that the hybrid datasets enhanced the F1-Score by 4.4%, from 0.4154 to 0.4329, compared to virtual images alone, and by 10%, from 0.4499 to 0.4990, compared to sole reliance on real image augmentation, underscoring its superiority. This investigation contributes to unmanned, automated quality inspection aligning with smart construction management, potentially bolstering productivity in the construction industry.

AeroDetectNet: a lightweight, high-precision network for enhanced detection of small objects in aerial remote sensing imagery

Object detection in remote sensing imagery exhibits difficulties due to complex backgrounds, diverse object scales, and intricate spatial context relationships. Motivated by the problems mentioned above, this paper introduces AeroDetectNet, a novel lightweight and high-precision object detection network custom-designed for aerial remote sensing scenarios, building upon the YOLOv7-tiny algorithm. It enhances performance through four key improvements: the normalized Wasserstein distance for consistent object size sensitivity, the Involution module for reduced background noise, a self-designed RCS-biformer module for better spatial context interpretation, and a self-designed WF-CoT SPPCSP feature pyramid for improved feature map weighting and context capture. Ablation studies conducted on a hybrid dataset composed of three open-source remote sensing datasets (including NWPU VHR-10 remote sensing images, RSOD remote sensing images, and VisDrone UAV images) have demonstrated the effectiveness of four improvements specifically for small-size object detection. Visualizations through Grad-CAM further demonstrate AeroDetectNet’s capacity to extract and focus on key object features. Upon individual testing across three open-source datasets, AeroDetectNet has successfully demonstrated its ability to identify objects in images with a smaller pixel area. Through experimental comparisons with other related studies, the AeroDetectNet achieved a competitive mAP while maintaining fewer model parameters, highlighting its highly accurate and lightweight properties.

ANALYZE-AD: A comparative analysis of novel AI approaches for early Alzheimer’s detection

Alzheimer’s disease, characterized by progressive and irreversible deterioration of cognitive functions, represents a significant health concern, particularly among older adults, as it stands as the foremost cause of dementia. Despite its debilitating nature, early detection of Alzheimer’s disease holds considerable advantages for affected individuals. This study investigates machine-learning methodologies for the early diagnosis of Alzheimer’s disease, utilizing datasets sourced from OASIS and ADNI. The initial classification methods consist of a 5-class ADNI classification and a 3-class OASIS classification. Three unique methodologies encompass binary-class inter-dataset models, which involve training on a single dataset and subsequently testing on another dataset for both ADNI and OASIS datasets. Additionally, a hybrid dataset model is also considered. The proposed methodology entails the concatenation of both datasets, followed by shuffling and subsequently conducting training and testing on the amalgamated dataset. The findings demonstrate impressive levels of accuracy, as Light Gradient Boosting Machine (LGBM) achieved a 99.63% accuracy rate for 5-class ADNI classification and a 95.75% accuracy rate by Multilayer Perceptron (MLP) for 3-class OASIS classification, both when hyperparameter tweaking was implemented. The K-nearest neighbor algorithm demonstrated exceptional performance, achieving an accuracy of 87.50% in ADNI-OASIS (2 Class) when utilizing the Select K Best method. The Gaussian Naive Bayes algorithm demonstrated exceptional performance in the OASIS-ADNI approach, attaining an accuracy of 77.97% using Chi-squared feature selection. The accuracy achieved by the Hybrid method, which utilized LGBM with hyperparameter optimization, was 99.21%. Furthermore, the utilization of Explainable AI approaches, particularly Lime, was implemented in order to augment the interpretability of the model.

Masked self‐supervised pre‐training model for EEG‐based emotion recognition

AbstractElectroencephalogram (EEG), as a tool capable of objectively recording brain electrical signals during emotional expression, has been extensively utilized. Current technology heavily relies on datasets, with its performance being limited by the size of the dataset and the accuracy of its annotations. At the same time, unsupervised learning and contrastive learning methods largely depend on the feature distribution within datasets, thus requiring training tailored to specific datasets for optimal results. However, the collection of EEG signals is influenced by factors such as equipment, settings, individuals, and experimental procedures, resulting in significant variability. Consequently, the effectiveness of models is heavily dependent on dataset collection efforts conducted under stringent objective conditions. To address these challenges, we introduce a novel approach: employing a self‐supervised pre‐training model, to process data across different datasets. This model is capable of operating effectively across multiple datasets. The model conducts self‐supervised pre‐training without the need for direct access to specific emotion category labels, enabling it to pre‐train and extract universally useful features without predefined downstream tasks. To tackle the issue of semantic expression confusion, we employed a masked prediction model that guides the model to generate richer semantic information through learning bidirectional feature combinations in sequence. Addressing challenges such as significant differences in data distribution, we introduced adaptive clustering techniques that manage by generating pseudo‐labels across multiple categories. The model is capable of enhancing the expression of hidden features in intermediate layers during the self‐supervised training process, enabling it to learn common hidden features across different datasets. This study, by constructing a hybrid dataset and conducting extensive experiments, demonstrated two key findings: (1) our model performs best on multiple evaluation metrics; (2) the model can effectively integrate critical features from different datasets, significantly enhancing the accuracy of emotion recognition.

Sentiment analysis of student feedback using attention-based RNN and transformer embedding

Sentiment analysis systems aim to assess people’s opinions across various domains by collecting and categorizing feedback and reviews. In our study, researchers put forward a sentiment analysis system that leverages three distinct embedding techniques: automatic, global vectors (GloVe) for word representation, and bidirectional encoder representations from transformers (BERT). This system features an attention layer, with the best model chosen through rigorous comparisons. In developing the sentiment analysis model, we employed a hybrid dataset comprising students’ feedback and comments. This dataset comprises 3,820 comments, including 2,773 from formal evaluations and 1,047 generated by ChatGPT and prompting engineering. Our main motivation for integrating generative AI was to balance both positive and negative comments. We also explored recurrent neural network (RNN), gated recurrent unit (GRU), long short-term memory (LSTM), and bidirectional long short-term memory (Bi-LSTM), with and without pre-trained GloVe embedding. These techniques produced F-scores ranging from 67% to 69%. On the other hand, the sentiment model based on BERT, particularly its KERAS implementation, achieved higher F-scores ranging from 83% to 87%. The Bi-LSTM architecture outperformed other models and the inclusion of an attention layer further enhanced the performance, resulting in F-scores of 89% and 88% from the Bi-LSTM-BERT sentiment models, respectively.

Detecting road signs and deformations for road safety in Senegal

Nowadays, road safety is a major concern for public authorities on national, secondary and classified roads in developing countries. This paper looks at the problem of non-compliance with road signs and the deteriorating state of roads, which are the cause of many accidents. Road users sometimes miss roadside signs because their attention is focused on the road. At other times, they regularly miss certain road markings or road deformations in unfamiliar areas, which is dangerous for travelers. Ideally, therefore, drivers should be able to be alerted to road signs or deformations from a distance without having to divert their attention. This would make it possible to avoid accidents, especially on national roads between regions of a country and between countries in sub-Saharan Africa.In this paper, we propose a model for the detection and recognition of road signs and deformations using deep learning. The dataset for this model will consist of a hybrid dataset that includes Kaggle data and data from AGEROUTE (Autonomous Agency for Works and Road Management) in Senegal. In addition, the proposed model is trained on two datasets; one for training 80% and the other for testing 20%. Experimental results showed that the proposed model achieved an accuracy of 95.34% for the datasets.

Experimental Study on Using Synthetic Images as a Portion of Training Dataset for Object Recognition in Construction Site

The application of Artificial Intelligence (AI) across various industries necessitates the acquisition of relevant environmental data and the implementation of AI recognition learning based on this data. However, the data available in real-world environments are limited and difficult to obtain. Construction sites represent dynamic and hazardous environments with a significant workforce, making data acquisition challenging and labor-intensive. To address these issues, this experimental study explored the potential of generating synthetic data to overcome the challenges of obtaining data from hazardous construction sites. Additionally, this research investigated the feasibility of hybrid dataset in securing construction-site data by creating synthetic data for scaffolding, which has a high incidence of falls but low object recognition rates due to its linear object characteristics. We generated a dataset by superimposing scaffolding objects, from which the backgrounds were removed, onto various construction site background images. Using this dataset, we produced a hybrid dataset to assess the feasibility of synthetic data for construction sites and to evaluate improvements in object recognition performance. By finding the optimal composition ratio with real data and conducting model training, the highest accuracy was achieved at an 8:2 ratio, with a construction object recognition accuracy of 0.886. Therefore, this study aims to reduce the risk and labor associated with direct data collection at construction sites through a hybrid dataset, achieving data generation at a low cost and high efficiency. By generating synthetic data to find the optimal ratio and constructing a hybrid dataset, this research demonstrates the potential to address the problems of data scarcity and data quality on construction sites. The improvement in recognition accuracy of the construction safety management system is anticipated, suggesting that the creation of synthetic data for constructing a hybrid dataset can reduce construction safety-accident issues.

Enhancing Moisture-Induced Defect Detection in Insulated Steel Pipes through Infrared Thermography and Hybrid Dataset

It is crucial to accurately detect moisture-induced defects in steel pipe insulation in order to combat corrosion under insulation (CUI). This study enhances the capabilities of infrared thermography (IRT) by integrating it with top-performing machine learning models renowned for their effectiveness in image segmentation tasks. A novel methodology was developed to enrich machine learning training, incorporating synthetic datasets generated via finite element method (FEM) simulations with experimental data. The performance of four advanced models—UNet, UNet++, DeepLabV3+, and FPN—was evaluated. These models demonstrated significant enhancements in defect detection capabilities, with notable improvements observed in FPN, which exhibited a mean intersection over union (IoU) increase from 0.78 to 0.94, a reduction in loss from 0.19 to 0.06, and an F1 score increase from 0.92 to 0.96 when trained on hybrid datasets compared to those trained solely on real data. The results highlight the benefits of integrating synthetic and experimental data, effectively overcoming the challenges of limited dataset sizes, and significantly improving the models’ accuracy and generalization capabilities in identifying defects. This approach marks a significant advancement in industrial maintenance and inspection, offering a precise, reliable, and scalable solution to managing the risks associated with CUI.

On the design of deep learning-based control algorithms for visually guided UAVs engaged in power tower inspection tasks.

This paper focuses on the design of Convolution Neural Networks to visually guide an autonomous Unmanned Aerial Vehicle required to inspect power towers. The network is required to precisely segment images taken by a camera mounted on a UAV in order to allow a motion module to generate collision-free and inspection-relevant manoeuvres of the UAV along different types of towers. The images segmentation process is particularly challenging not only because of the different structures of the towers but also because of the enormous variability of the background, which can vary from the uniform blue of the sky to the multi-colour complexity of a rural, forest, or urban area. To be able to train networks that are robust enough to deal with the task variability, without incurring into a labour-intensive and costly annotation process of physical-world images, we have carried out a comparative study in which we evaluate the performances of networks trained either with synthetic images (i.e., the synthetic dataset), physical-world images (i.e., the physical-world dataset), or a combination of these two types of images (i.e., the hybrid dataset). The network used is an attention-based U-NET. The synthetic images are created using photogrammetry, to accurately model power towers, and simulated environments modelling a UAV during inspection of different power towers in different settings. Our findings reveal that the network trained on the hybrid dataset outperforms the networks trained with the synthetic and the physical-world image datasets. Most notably, the networks trained with the hybrid dataset demonstrates a superior performance on multiples evaluation metrics related to the image-segmentation task. This suggests that, the combination of synthetic and physical-world images represents the best trade-off to minimise the costs related to capturing and annotating physical-world images, and to maximise the task performances. Moreover, the results of our study demonstrate the potential of photogrammetry in creating effective training datasets to design networks to automate the precise movement of visually-guided UAVs.

Bayesian data-driven framework for structural health monitoring of composite structures under limited experimental data

Ultrasonic-guided waves can be used to monitor the health of thin-walled structures. However, the run of experimental damage tests on materials like carbon fiber-reinforced plastics can be impractical and costly. Instead, numerical models can be used to create hybrid datasets to train machine learning algorithms, integrating data from numerical and experimental tests. This paper presents a Bayesian-driven framework to compensate for limited experimental data regarding Lamb wave propagation in composite plates. Using Bayesian inference, the framework updates a numerical finite element model, considering observed uncertainties by sampling posterior probability density functions for input parameters using Markov–Chain Monte Carlo simulations with the Metropolis-Hastings algorithm. A neural network surrogate model speeds-up these simulations, leading to a model that replicates the uncertain experimental setup. This model then generates data to augment true experimental data. Finally, a one-dimensional convolutional neural network is trained on a three different datasets to analyze Lamb wave signals and assess damage. Comparing training strategies shows the hybrid dataset augmented by samples generated by the updated FE model gives the most accurate damage size predictions.

An auto‐configurable machine learning framework to optimize and predict catalysts for CO2 to light olefins process

AbstractThis study proposed an auto‐configurable machine learning framework based on the differential evolution algorithm (AutoML‐DE) driven by hybrid data for the screening and discovery of promising CO2 to light olefins (CO2TLO) catalysts candidates. The hybrid dataset comprises 532 experimental data from the literature and 296 simulation data. Results show that the AutoML‐DE model with extreme gradient boosting algorithms demonstrated superior performance for predicting the conversion ratio of CO2 and selectivity of light olefins (average R2 &gt; 0.86). After identifying the input feature with the most significant impact on the output feature, the optimal AutoML‐DE model integrated with the genetic algorithm is applied to multiobjective optimization, sensitivity analysis, and prediction of new CO2TLO catalysts. The optimized Cu‐Zn‐Al/SAPO‐34 catalyst has the highest catalytic performance among the reported CO2TLO catalysts. Moreover, five new CO2TLO catalysts with higher yields are successfully predicted. However, the performance of these catalysts should be further verified by experiment.

Stroke-volume-allocation model enabling wearable sensors for vascular age and cardiovascular disease assessment

Frequent and unobtrusive monitoring of cardiovascular conditions with consumer electronics is a widely pursued goal, since it provides the most economic and effective way of preventing and managing cardiovascular diseases (CVDs) ─ the leading causes of death worldwide. However, most current wearable and flexible devices can only support the measurement of one or two types of vital signs, such as heart rate and blood oxygen level, due to the lack of physiological models to link the measured signals to cardiovascular conditions. Here, we report a stroke-volume allocation (SVA) model to quantify the cushioning function of arteries and empower nearly all existing cardiac sensors with new functions, including arterial stiffness evaluation, dynamic blood pressure tracking and classification of CVD-related heart damage. Large-scale clinical data testing involving a hybrid dataset taken from 6 hospitals/research institutes (9 open databases and 4 self-built databases from 878 subjects in total) and diverse measurement approaches was carried out to validate the SVA model. The results show that the SVA-based parameters correlate well with the gold-standard measurements in arterial stiffness and blood pressure and outperform the commonly used vital sign (e.g., blood pressure) alone in detecting abnormalities in cardiovascular systems.

Innovative Point Cloud Segmentation of 3D Light Steel Framing System through Synthetic BIM and Mixed Reality Data: Advancing Construction Monitoring

In recent years, mixed reality (MR) technology has gained popularity in construction management due to its real-time visualisation capability to facilitate on-site decision-making tasks. The semantic segmentation of building components provides an attractive solution towards digital construction monitoring, reducing workloads through automation techniques. Nevertheless, data shortages remain an issue in maximizing the performance potential of deep learning segmentation methods. The primary aim of this study is to address this issue through synthetic data generation using Building Information Modelling (BIM) models. This study presents a point-cloud-based deep learning segmentation approach to a 3D light steel framing (LSF) system through synthetic BIM models and as-built data captured using MR headsets. A standardisation workflow between BIM and MR models was introduced to enable seamless data exchange across both domains. A total of five different experiments were set up to identify the benefits of synthetic BIM data in supplementing actual as-built data for model training. The results showed that the average testing accuracy using solely as-built data stood at 82.88%. Meanwhile, the introduction of synthetic BIM data into the training dataset led to an improved testing accuracy of 86.15%. A hybrid dataset also enabled the model to segment both the BIM and as-built data captured using an MR headset at an average accuracy of 79.55%. These findings indicate that synthetic BIM data have the potential to supplement actual data, reducing the costs associated with data acquisition. In addition, this study demonstrates that deep learning has the potential to automate construction monitoring tasks, aiding in the digitization of the construction industry.

Machine learning traction force maps for contractile cell monolayers

Machine learning offers immense potential as a transformative tool capable of reshaping optical microscopy and quantitative modeling in cell biology. Here we exemplify this potential through the development of a generative adversarial network (GAN) designed to comprehend and predict cell traction force maps. Empowered by a hybrid dataset from traction force microscopy (TFM) and phase-field modeling (PFM), the GAN learns the intricacies of the traction force maps of contractile cells in complex chemomechanical environments, with the sole input being the phase-contrast images of the cells. The trained GAN accurately predicts collective durotaxis by leveraging the learned asymmetric traction force maps, while also unveiling the concealed correlation between substrate stiffness and cell contractility arising from mechanotransduction. Remarkably, despite its foundation in epithelial cell data, our image-learning algorithm can be extended to other contractile cell types by adjusting a single scaling factor. Our approach underscores the potential of synergizing force microscopies and biophysical models with image-based learning, thus catalyzing data-driven scientific revelations in cell mechanobiology.