Efficient extraction of deep image features using convolutional neural network (CNN) for applications in detecting and analysing complex food matrices

Information extraction of UV-NIR spectral data in waste water based on Large Language Model

In recent years, with the rise of various machine learning methods, the Ultraviolet and Near Infrared (UV-NIR) spectral analysis has been impressive in the determination of intricate systems. However, the UV-NIR spectral analysis based on traditional machine learning requires independent training with tedious parameter tuning for different samples or tasks. As a result, training a high-quality model is often complicated and time-consuming. Large language model (LLM) is one of the cutting-edge achievements in deep learning, with the parameter size of the order of billion. LLM can extract abstract information from input and use it effectively. Even without any additional training, using only simple natural language prompts, LLM can accomplish tasks that have never been seen before in completely new domains. We look forward to utilizing this capability in spectral analysis to reduce the time-consuming and operational difficulties. In this study, we used UV-NIR spectral analysis to predict the concentration of Chemical Oxygen Demand (COD) in three different water samples, including a complex wastewater. By extracting the characteristic bands in the spectrum, we input them into LLM for concentration prediction. We compared the COD prediction results of different models on water samples and discussed the effects of different experiments setting on LLM. The results show that even with brief prompts, the prediction of LLM in wastewater achieved the best performance, with R2 and RMSE equal to 0.931 and 10.966, which exceed the best results of traditional models, where R2 and RMSE correspond to 0.920 and 11.854. This result indicates that LLM, with simpler operation and less time-consuming, has ability to approach or even surpass traditional machine learning models in UV-NIR spectral analysis. In conclusion, our study proposed a new method for the UV-NIR spectral analysis based on LLM and preliminary demonstrated the potential of LLM for application.

Integrating Vision-Language Models for Accelerated High-Throughput Nutrition Screening

Addressing the critical need for swift and precise nutritional profiling in healthcare and in food industry, this study pioneers the integration of vision-language models (VLMs) with chemical analysis techniques. A cutting-edge VLM is unveiled, utilizing the expansive UMDFood-90k database, to significantly improve the speed and accuracy of nutrient estimation processes. Demonstrating a macro-AUCROC of 0.921 for lipid quantification, the model exhibits less than 10% variance compared to traditional chemical analyses for over 82% of the analyzed food items. This innovative approach not only accelerates nutritional screening by 36.9% when tested amongst students but also sets a new benchmark in the precision of nutritional data compilation. This research marks a substantial leap forward in food science, employing a blend of advanced computational models and chemical validation to offer a rapid, high-throughput solution for nutritional analysis.

The Application of Multi-Parameter Multi-Modal Technology Integrating Biological Sensors and Artificial Intelligence in the Rapid Detection of Food Contaminants

Against the backdrop of continuous socio-economic development, there is a growing concern among people about food quality and safety. Individuals are increasingly realizing the critical importance of healthy eating for bodily health; hence the continuous rise in demand for detecting food pollution. Simultaneously, the rapid expansion of global food trade has made people's pursuit of high-quality food more urgent. However, traditional methods of food analysis have certain limitations, mainly manifested in the high degree of reliance on personal subjective judgment for assessing food quality. In this context, the emergence of artificial intelligence and biosensors has provided new possibilities for the evaluation of food quality. This paper proposes a comprehensive approach that involves aggregating data relevant to food quality indices and developing corresponding evaluation models to highlight the effectiveness and comprehensiveness of artificial intelligence and biosensors in food quality evaluation. The potential prospects and challenges of this method in the field of food safety are comprehensively discussed, aiming to provide valuable references for future research and practice.

Array-optimized artificial olfactory sensor enabling cost-effective and non-destructive detection of mycotoxin-contaminated maize

The MobileNetV3-based improved sine-cosine algorithm (ISCA-MobileNetV3) was combined with an artificial olfactory sensor (AOS) to address the redundancy in olfactory arrays, thereby achieving low-cost and high-precision detection of mycotoxin-contaminated maize. Specifically, volatile organic compounds of maize interacted with unoptimized AOS containing eight porphyrins and eight dye-attached nanocomposites to obtain the scent fingerprints for constructing the initial data set. The optimal decision model was MobileNetV3, with more than 98.5% classification accuracy, and its output training loss would be input into the optimizer ISCA. Remarkably, the number of olfactory arrays was reduced from 16 to 6 by ISCA-MobileNetV3 with about a 1% decrease in classification accuracy. Additionally, the developed system showed that each online evaluation was less than one second on average, demonstrating outstanding real-time performance for ensuring food safety. Therefore, AOS combined with ISCA-MobileNetV3 will encourage the development of an affordable and on-site platform for maize quality detection.

Advanced Image Analytics for Mobile Robot-Based Condition Monitoring in Hazardous Environments: A Comprehensive Thermal Defect Processing Framework

In hazardous environments like mining sites, mobile inspection robots play a crucial role in condition monitoring (CM) tasks, particularly by collecting various kinds of data, such as images. However, the sheer volume of collected image samples and existing noise pose challenges in processing and visualizing thermal anomalies. Recognizing these challenges, our study addresses the limitations of industrial big data analytics for mobile robot-generated image data. We present a novel, fully integrated approach involving a dimension reduction procedure. This includes a semantic segmentation technique utilizing the pre-trained VGG16 CNN architecture for feature selection, followed by random forest (RF) and extreme gradient boosting (XGBoost) classifiers for the prediction of the pixel class labels. We also explore unsupervised learning using the PCA-K-means method for dimension reduction and classification of unlabeled thermal defects based on anomaly severity. Our comprehensive methodology aims to efficiently handle image-based CM tasks in hazardous environments. To validate its practicality, we applied our approach in a real-world scenario, and the results confirm its robust performance in processing and visualizing thermal data collected by mobile inspection robots. This affirms the effectiveness of our methodology in enhancing the overall performance of CM processes.

A convolutional neural network with image and numerical data to improve farming of edible crickets as a source of food-A decision support system

Crickets (Gryllus bimaculatus) produce sounds as a natural means to communicate and convey various behaviors and activities, including mating, feeding, aggression, distress, and more. These vocalizations are intricately linked to prevailing environmental conditions such as temperature and humidity. By accurately monitoring, identifying, and appropriately addressing these behaviors and activities, the farming and production of crickets can be enhanced. This research implemented a decision support system that leverages machine learning (ML) algorithms to decode and classify cricket songs, along with their associated key weather variables (temperature and humidity). Videos capturing cricket behavior and weather variables were recorded. From these videos, sound signals were extracted and classified such as calling, aggression, and courtship. Numerical and image features were extracted from the sound signals and combined with the weather variables. The extracted numerical features, i.e., Mel-Frequency Cepstral Coefficients (MFCC), Linear Frequency Cepstral Coefficients, and chroma, were used to train shallow (support vector machine, k-nearest neighbors, and random forest (RF)) ML algorithms. While image features, i.e., spectrograms, were used to train different state-of-the-art deep ML models, i,e., convolutional neural network architectures (ResNet152V2, VGG16, and EfficientNetB4). In the deep ML category, ResNet152V2 had the best accuracy of 99.42%. The RF algorithm had the best accuracy of 95.63% in the shallow ML category when trained with a combination of MFCC+chroma and after feature selection. In descending order of importance, the top 6 ranked features in the RF algorithm were, namely humidity, temperature, C#, mfcc11, mfcc10, and D. From the selected features, it is notable that temperature and humidity are necessary for growth and metabolic activities in insects. Moreover, the songs produced by certain cricket species naturally align to musical tones such as C# and D as ranked by the algorithm. Using this knowledge, a decision support system was built to guide farmers about the optimal temperature and humidity ranges and interpret the songs (calling, aggression, and courtship) in relation to weather variables. With this information, farmers can put in place suitable measures such as temperature regulation, humidity control, addressing aggressors, and other relevant interventions to minimize or eliminate losses and enhance cricket production.

Detection of peach soluble solids based on near-infrared spectroscopy with High Order Spatial Interaction network

BACKGROUND

Due to the scalability of deep learning technology, researchers have applied it to the non-destructive testing of peach internal quality. In addition, the soluble solids content (SSC) is an important internal quality indicator that determines the quality of peaches. Peaches with high SSC have a sweeter taste and better texture, making them popular in the market. Therefore, SSC is an important indicator for measuring peach internal quality and making harvesting decisions.

RESULTS

This article presents the High Order Spatial Interaction Network (HOSINet), which combines the Position Attention Module (PAM) and Channel Attention Module (CAM). Additionally, a feature wavelength selection algorithm similar to the Group-based Clustering Subspace Representation (GCSR-C) is used to establish the Position and Channel Attention Module-High Order Spatial Interaction (PC-HOSI) model for peach SSC prediction. The accuracy of this model is compared with traditional machine learning and traditional deep learning models. Finally, the permutation algorithm is combined with deep learning models to visually evaluate the importance of feature wavelengths. Increasing the order of the PC-HOSI model enhances its ability to learn spatial correlations in the dataset, thus improving its predictive performance.

CONCLUSION

The optimal model, PC-HOSI model, performed well with an order of 3 (PC-HOSI-3), with a root mean square error of 0.421 °Brix and a coefficient of determination of 0.864. Compared with traditional machine learning and deep learning algorithms, the coefficient of determination for the prediction set was improved by 0.07 and 0.39, respectively. The permutation algorithm also provided interpretability analysis for the predictions of the deep learning model, offering insights into the importance of spectral bands. These results contribute to the accurate prediction of SSC in peaches and support research on interpretability of neural network models for prediction. © 2024 Society of Chemical Industry.

Non-destructive prediction of TVB-N using color-texture features of UV-induced fluorescence image for freeze-thaw treated frozen-whole-round tilapia

BACKGROUND

The investigation of UV-induced fluorescence imaging coupled with machine learning was conducted to non-destructively detect the total volatile basic nitrogen (TVB-N) of frozen-whole-round tilapia (FWRT) during freezing and thawing. The UV-induced fluorescence images of FWRT at the wavelength of 365 nm were acquired by self-developed fluorescence image acquisition system. In total, 169 color and texture features based on RGB, hue-saturation-intensity and L*a*b* color spaces and gray level co-occurrence matrix were extracted, respectively. Successive projections algorithm (SPA) was employed to select the optimal 16 features to achieve feature dimension reduction modeling. With full and extracted features as input, the models of partial least squares regression (PLSR), least-squares support vector machine (LSSVM) and convolutional neural network (CNN) were established for TVB-N prediction.

RESULTS

Results indicated that the full features-based CNN performed better than SPA based prediction models (SPA-PLSR and SPA-LSSVM). The CNN model was determined to be the optimal with an RP2 value of 0.9779, RMSEP value of 1.1502 × 10-2  g N kg-1 and RPD value of 6.721 for TVB-N content predictiin.

CONCLUSION

The CNN method based on UV fluorescence imaging technology has potential for quality and safety detection of FWRT. © 2023 Society of Chemical Industry.

Efficiency of Identification of Blackcurrant Powders Using Classifier Ensembles

In the modern times of technological development, it is important to select adequate methods to support various food and industrial problems, including innovative techniques with the help of artificial intelligence (AI). Effective analysis and the speed of algorithm implementation are key points in assessing the quality of food products. Non-invasive solutions are being sought to achieve high accuracy in the classification and evaluation of various food products. This paper presents various machine learning algorithm architectures to evaluate the efficiency of identifying blackcurrant powders (i.e., blackcurrant concentrate with a density of 67 °Brix and a color coefficient of 2.352 (E520/E420) in combination with the selected carrier) based on information encoded in microscopic images acquired via scanning electron microscopy (SEM). Recognition of blackcurrant powders was performed using texture feature extraction from images aided by the gray-level co-occurrence matrix (GLCM). It was evaluated for quality using individual single classifiers and a metaclassifier based on metrics such as accuracy, precision, recall, and F1-score. The research showed that the metaclassifier, as well as a single random forest (RF) classifier most effectively identified blackcurrant powders based on image texture features. This indicates that ensembles of classifiers in machine learning is an alternative approach to demonstrate better performance than the existing traditional solutions with single neural models. In the future, such solutions could be an important tool to support the assessment of the quality of food products in real time. Moreover, ensembles of classifiers can be used for faster analysis to determine the selection of an adequate machine learning algorithm for a given problem.

DeepTrayMeal: Automatic dietary assessment for Chinese tray meals based on deep learning

Tray meal is a popular way of eating in China, and tray-based automatic dietary assessment is important for public health. Relevant research is lacking because public tray meal datasets and suitable methods are unavailable. In this study, we established and published the first Chinese tray meal dataset, the ChinaLunchTray-99. We collected real-world 1185 tray meal images, covering 99 dish categories with corresponding manually annotated bounding box and category-level labels. We developed a new framework for automatic dietary assessment, which consists of dish image recognition, volume estimation and nutrition mapping. First, we demonstrated a tray meal detection model considering feature extraction, anchor scales, and loss function, resulting in a high mean Average Precision of 92.13%. Second, we proposed an automatic method to estimate volume via detection results and tray's information. Finally, nutrients were mapped from the estimated volume. Our research can promote applications of automatic dietary assessment for Chinese tray meals.

Development and comparison of parallel reaction monitoring and data-independent acquisition methods for quantitative analysis of hydrophilic compounds in white tea

In the present work, parallel reaction monitoring (PRM) and data-independent acquisition (DIA) methods were developed for the accurate quantitation of amino acids, alkaloids nucleosides and nucleotides in tea. The quality peaks were significantly enhanced by optimizing the LC elution procedure, HCD voltage, MS resolution, and scanning event. Both methods were validated with good liner linearity (0.004-200 μg/mL), LODs (0.001-0.309 μg/mL for PRM and 0.001-0.564 μg/mL for DIA). Applied to white tea sample, the contents of these hydrophilic compounds were range from 34,655.39 to 70,586.14 mg/kg, and caffeine (32,529.02 mg/kg) and theanine (5483.46 mg/kg) were determined as the most abundant ones. Based on the quantitation data set, the white tea samples from Puer, Lincang and Xishuangbanna were clearly discriminated using multivariate data analysis. The results of the present works show that PRM and DIA have great potential in quantitative analysis of multiple hydrophilic compounds in food samples.

Integrative deep learning framework predicts lipidomics-based investigation of preservatives on meat nutritional biomarkers and metabolic pathways

Preservatives are added as antimicrobial agents to extend the shelf life of meat. Adding preservatives to meat products can affect their flavor and nutrition. This review clarifies the effects of preservatives on metabolic pathways and network molecular transformations in meat products based on lipidomics, metabolomics and proteomics analyses. Preservatives change the nutrient content of meat products via altering ionic strength and pH to influence enzyme activity. Ionic strength in salt triggers muscle triglyceride hydrolysis by causing phosphorylation and lipid droplet splitting in adipose tissue hormone-sensitive lipase and triglyceride lipase. DisoLipPred exploiting deep recurrent networks and transfer learning can predict the lipid binding trend of each amino acid in the disordered region of input protein sequences, which could provide omics analyses of biomarkers metabolic pathways in meat products. While conventional meat quality assessment tools are unable to elucidate the intrinsic mechanisms and pathways of variables in the influences of preservatives on the quality of meat products, the promising application of omics techniques in food analysis and discovery through multimodal learning prediction algorithms of neural networks (e.g., deep neural network, convolutional neural network, artificial neural network) will drive the meat industry to develop new strategies for food spoilage prevention and control.

Mixed pesticide recognition based on three-dimensional fluorescence spectroscopy and a convolutional neural network

Kasugamycin, spinosad, and lambda-cyhalothrin are common organic pesticides that are widely used to control and prevent diseases and pests in fruits and vegetables. However, the unreasonable use of pesticides will cause great harm to the natural environment and human health. Pesticides often exist in the form of mixtures in nature. Establishing recognition models for mixed pesticides in large-scale sample testing can provide guidance for further precise analysis and reduce resource waste and time. Therefore, finding a fast and effective identification method for mixed pesticides is of great significance. This paper applies three-dimensional fluorescence spectroscopy to detect mixed pesticides and introduces a convolutional neural network (CNN) model structure based on an improved LeNet-5 to classify mixed pesticides. The input part of the model corresponds to fluorescence spectrum data at excitation wavelengths of 250-306 nm and emission wavelengths of 300-450 nm, and the mixed pesticides are divided into three categories. The research results show that when the learning rate is set to 1 and the number of iterations is 300, the CNN classification model has ideal performance (with a recognition accuracy of 100%) and is superior to the performance of the support vector machine method. This paper provides a certain methodological basis for the rapid identification of mixed pesticides.

Design of an Artificial Intelligence of Things-Based Sesame Oil Evaluator for Quality Assessment Using Gas Sensors and Deep Learning Mechanisms

Traditional oil quality measurement is mostly based on chemical indicators such as acid value, peroxide value, and p-anisidine value. This process requires specialized knowledge and involves complex steps. Hence, this study designs and proposes a Sesame Oil Quality Assessment Service Platform, which is composed of an Intelligent Sesame Oil Evaluator (ISO Evaluator) and a Cloud Service Platform. Users can quickly assess the quality of sesame oil using this platform. The ISO Evaluator employs Artificial Intelligence of Things (AIoT) sensors to detect changes in volatile gases and the color of the oil during storage. It utilizes deep learning mechanisms, including Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) to determine and evaluate the quality of the sesame oil. Evaluation results demonstrate that the linear discriminant analysis (LDA) value is 95.13. The MQ2, MQ3, MQ4, MQ7, and MQ8 sensors have a positive correlation. The CNN combined with an ANN model achieves a Mean Absolute Percentage Error (MAPE) of 8.1820% for predicting oil quality, while the LSTM model predicts future variations in oil quality indicators with a MAPE of 0.44%. Finally, the designed Sesame Oil Quality Assessment Service Platform effectively addresses issues related to digitization, quality measurement, supply quality observation, and scalability.

A Spatial-Spectral Classification Method Based on Deep Learning for Controlling Pelagic Fish Landings in Chile

Fishing has provided mankind with a protein-rich source of food and labor, allowing for the development of an important industry, which has led to the overexploitation of most targeted fish species. The sustainable management of these natural resources requires effective control of fish landings and, therefore, an accurate calculation of fishing quotas. This work proposes a deep learning-based spatial-spectral method to classify five pelagic species of interest for the Chilean fishing industry, including the targeted Engraulis ringens, Merluccius gayi, and Strangomera bentincki and non-targeted Normanichthtys crockeri and Stromateus stellatus fish species. This proof-of-concept method is composed of two channels of a convolutional neural network (CNN) architecture that processes the Red-Green-Blue (RGB) images and the visible and near-infrared (VIS-NIR) reflectance spectra of each species. The classification results of the CNN model achieved over 94% in all performance metrics, outperforming other state-of-the-art techniques. These results support the potential use of the proposed method to automatically monitor fish landings and, therefore, ensure compliance with the established fishing quotas.

Rapid identification of counterfeited beef using deep learning-aided spectroscopy: Detecting colourant and curing agent adulteration

The adulteration of meat products using colourants and curing agents has heightened concerns over food safety, thereby necessitating the development of advanced detection methods. This study introduces a deep-learning-based spectroscopic method for swiftly identifying counterfeit beef altered to appear fresh. The experiment involved 60 beef samples, half of which were artificially adulterated using a colouring solution. Despite meticulous analysis of the beef's colour attributes, no significant differences were observed between the fresh and adulterated samples. However, our method, utilising a 344-1040 nm spectral range, achieved a classification accuracy of 98.84%. To enhance practicality, we employed gradient-weighted class activation mapping and identified the 580-600 nm range as particularly influential for classification. Remarkably, even when we narrowed the input to the model to this spectral range, a high level of classification accuracy was maintained. To further validate the model's robustness and generalisability, we allocated 70 beef samples to an external validation set. Comparative performance analysis revealed that our model outperformed traditional machine learning algorithms, such as SVM and logistic regression, by 9.3% and 28.4%, respectively. Overall, this study offers invaluable insights for detecting counterfeited beef, thereby contributing to the preservation of meat product quality and integrity within the food industry.

Classification of benign and malignant parotid tumors based on CT images combined with stack generalization model

Parotid tumors are among the most prevalent tumors in otolaryngology, and malignant parotid tumors are one of the main causes of facial paralysis in patients. Currently, the main diagnostic modality for parotid tumors is computed tomography, which relies mainly on the subjective judgment of clinicians and leads to practical problems such as high workloads. Therefore, to assist physicians in solving the preoperative classification problem, a stacked generalization model is proposed for the automated classification of parotid tumor images. A ResNet50 pretrained model is used for feature extraction. The first layer of the adopted stacked generalization model consists of multiple weak learners, and the results of the weak learners are integrated as input data in a meta-classifier in the second layer. The output results of the meta-classifier are the final classification results. The classification accuracy of the stacked generalization model reaches 91%. Comparing the classification results under different classifiers, the stacked generalization model used in this study can identify benign and malignant tumors in the parotid gland effectively, thus relieving physicians of tedious work pressure.

Machine learning applications for multi-source data of edible crops: A review of current trends and future prospects

The quality and safety of edible crops are key links inseparable from human health and nutrition. In the era of rapid development of artificial intelligence, using it to mine multi-source information on edible crops provides new opportunities for industrial development and market supervision of edible crops. This review comprehensively summarized the applications of multi-source data combined with machine learning in the quality evaluation of edible crops. Multi-source data can provide more comprehensive and rich information from a single data source, as it can integrate different data information. Supervised and unsupervised machine learning is applied to data analysis to achieve different requirements for the quality evaluation of edible crops. Emphasized the advantages and disadvantages of techniques and analysis methods, the problems that need to be overcome, and promising development directions were proposed. To monitor the market in real-time, the quality evaluation methods of edible crops must be innovated.

A novel model based on a 1D-ResCNN and transfer learning for processing EEG attenuation

EEG signals are valuable signals in clinical medicine, brain research, and the study of neurological illnesses. However, EEG signal attenuation may occur at any time from signal generation through BCI device acquisition due to defects in the brain-computer interface (BCI) devices, restrictions in the dynamic network, and individual variations across the subjects. The attenuation of EEG data will alter the data distribution and lead to information fuzziness, substantially influencing subsequent EEG research. A model based on one-dimensional residual convolutional neural networks (1D-ResCNN) and transfer learning is proposed in this article to reduce the negative impacts of EEG attenuation. An end-to-end manner maps an attenuated EEG signal to a normal EEG signal. The structure employs a multi-level residual connection structure with varying weight coefficients, transferring characteristics from the bottom to the top of the convolutional neural network, enhancing feature learning. In addition, we initialize the subsequent denoising model using the transfer learning method. The combination of these two networks can well solve the attenuation problem of EEG signals. Experiments are carried out using the EEG-denoisenet data set. According to the findings, the model can yield a clear waveform with a decent SNR and RRMSE value.

A Vis/NIR spectra-based approach for identifying bananas infected with Colletotrichum musae

Introduction

Anthracnose of banana caused by Colletotrichum species is one of the most serious post-harvest diseases, which can cause significant yield losses. Clarifying the infection mechanism of the fungi using non-destructive methods is crucial for timely discriminating infected bananas and taking preventive and control measures.

Methods

This study presented an approach for tracking growth and identifying different infection stages of the C. musae in bananas using Vis/NIR spectroscopy. A total of 330 banana reflectance spectra were collected over ten consecutive days after inoculation, with a sampling rate of 24 h. The four-class and five-class discriminant patterns were designed to examine the capability of NIR spectra in discriminating bananas infected at different levels (control, acceptable, moldy, and highly moldy), and different time at early stage (control and days 1-4). Three traditional feature extraction methods, i.e. PC loading coefficient (PCA), competitive adaptive reweighted sampling (CARS) and successive projections algorithm (SPA), combining with two machine learning methods, i.e. partial least squares discriminant analysis (PLSDA) and support vector machine (SVM), were employed to build discriminant models. One-dimensional convolutional neural network (1D-CNN) without manually extracted feature parameters was also introduced for comparison.

Results

The PCA-SVM and·SPA-SVM models had good performance with identification accuracies of 93.98% and 91.57%, 94.47% and 89.47% in validation sets for the four- and five-class patterns, respectively. While the 1D-CNN models performed the best, achieving an accuracy of 95.18% and 97.37% for identifying infected bananas at different levels and time, respectively.

Discussion

These results indicate the feasibility of identifying banana fruit infected with C. musae using Vis/NIR spectra, and the resolution can be accurate to one day.

Vision-based food nutrition estimation via RGB-D fusion network

With the development of deep learning technology, vision-based food nutrition estimation is gradually entering the public view for its advantage in accuracy and efficiency. In this paper, we designed one RGB-D fusion network, which integrated multimodal feature fusion (MMFF) and multi-scale fusion for visioin-based nutrition assessment. MMFF performed effective feature fusion by a balanced feature pyramid and convolutional block attention module. Multi-scale fusion fused different resolution features through feature pyramid network. Both enhanced feature representation to improve the performance of the model. Compared with state-of-the-art methods, the mean value of the percentage mean absolute error (PMAE) for our method reached 18.5%. The PMAE of calories and mass reached 15.0% and 10.8% via the RGB-D fusion network, improved by 3.8% and 8.1%, respectively. Furthermore, this study visualized the estimation results of four nutrients and verified the validity of the method. This research contributed to the development of automated food nutrient analysis (Code and models can be found at http://123.57.42.89/codes/RGB-DNet/nutrition.html).

Distinguishing pericarpium citri reticulatae of different origins using terahertz time-domain spectroscopy combined with convolutional neural networks

The geographical indication of pericarpium citri reticulatae (PCR) is very important in grading the quality and price of PCRs. Therefore, terahertz time-domain spectroscopy (THz-TDS) technology combined with convolutional neural networks (CNN) was proposed to distinguish PCRs of different origins without damage in this study. The one-dimensional CNN (1D-CNN) model with an accuracy of 82.99% based on spectral data processed with SNV was established. The two-dimensional image features were transformed from unprocessed spectral data using the gramian angular field (GAF), the Markov transition field (MTF) and the recurrence plot (RP), which were used to build a two-dimensional CNN (2D-CNN) model with an accuracy of 78.33%. Further, the CNN models with different fusion methods were developed for fusing spectra data and image data. In addition, the adding spectra and images based on the CNN (Add-CNN) model with an accuracy of 86.17% performed better. Eventually, the Add-CNN model based on ten frequencies extracted using permutation importance (PI) achieved the identification of PCRs from different origins. Overall, the current study would provide a new method for identifying PCRs of different origins, which was expected to be used for the traceability of PCRs products.

Multi-task convolutional neural network for simultaneous monitoring of lipid and protein oxidative damage in frozen-thawed pork using hyperspectral imaging

Lipid and protein oxidation are the main causes of meat deterioration during freezing. Traditional methods using hyperspectral imaging (HSI) need to train multiple independent models to predict multiple attributes, which is complex and time-consuming. In this study, a multi-task convolutional neural network (CNN) model was developed for visible near-infrared HSI data (400-1002 nm) of 240 pork samples treated with different freeze-thaw cycles (0-9 cycles) to evaluate the feasibility of simultaneously monitoring lipid oxidation (thiobarbituric acid reactive substance content) and protein oxidation (carbonyl content) in pork. The performance of the commonly used partial least squares regression (PLSR) model based on the spectra after pre-processing (Standard normal variate, Savitzky-Golay derivative, and Savitzky-Golay smoothing) and feature selection (Regression coefficients) and single-output CNN model was compared. The results showed that the multi-task CNN model achieved the optimal prediction accuracies for lipid oxidation (R²p = 0.9724, RMSEP = 0.0227, and RPD = 5.2579) and protein oxidation (R²p = 0.9602, RMSEP = 0.0702, and RPD = 4.6668). In final, the changes of lipid and protein oxidation of pork in different freeze-thaw cycles were successfully visualized. In conclusion, the combination of HSI and multi-task CNN method shows the potential of end-to-end prediction of pork oxidative damage. This study provides a new, convenient and automated technique for meat quality detection in the food industry.

Quality assessment of traditional Chinese medicine based on data fusion combined with machine learning: A review

The authenticity and quality of traditional Chinese medicine (TCM) directly impact clinical efficacy and safety. Quality assessment of traditional Chinese medicine (QATCM) is a global concern due to increased demand and shortage of resources. Recently, modern analytical technologies have been extensively investigated and utilized to analyze the chemical composition of TCM. However, a single analytical technique has some limitations, and judging the quality of TCM only from the characteristics of the components is not enough to reflect the overall view of TCM. Thus, the development of multi-source information fusion technology and machine learning (ML) has further improved QATCM. Data information from different analytical instruments can better understand the connection between herbal samples from multiple aspects. This review focuses on the use of data fusion (DF) and ML in QATCM, including chromatography, spectroscopy, and other electronic sensors. The common data structures and DF strategies are introduced, followed by ML methods, including fast-growing deep learning. Finally, DF strategies combined with ML methods are discussed and illustrated for research on applications such as source identification, species identification, and content prediction in TCM. This review demonstrates the validity and accuracy of QATCM-based DF and ML strategies and provides a reference for developing and applying QATCM methods.

Artificial Intelligence Techniques Used to Extract Relevant Information from Complex Social Networks

Social networks constitute an almost endless source of social behavior information. In fact, sometimes the amount of information is so large that the task to extract meaningful information becomes impossible due to temporal constrictions. We developed an artificial-intelligence-based method that reduces the calculation time several orders of magnitude when conveniently trained. We exemplify the problem by extracting data freely available in a commonly used social network, Twitter, building up a complex network that describes the online activity patterns of society. These networks are composed of a huge number of nodes and an even larger number of connections, making extremely difficult to extract meaningful data that summarizes and/or describes behaviors. Each network is then rendered into an image and later analyzed using an AI method based on Convolutional Neural Networks to extract the structural information.

Artificial Intelligence in Food Safety: A Decade Review and Bibliometric Analysis

Artificial Intelligence (AI) technologies have been powerful solutions used to improve food yield, quality, and nutrition, increase safety and traceability while decreasing resource consumption, and eliminate food waste. Compared with several qualitative reviews on AI in food safety, we conducted an in-depth quantitative and systematic review based on the Core Collection database of WoS (Web of Science). To discover the historical trajectory and identify future trends, we analysed the literature concerning AI technologies in food safety from 2012 to 2022 by CiteSpace. In this review, we used bibliometric methods to describe the development of AI in food safety, including performance analysis, science mapping, and network analysis by CiteSpace. Among the 1855 selected articles, China and the United States contributed the most literature, and the Chinese Academy of Sciences released the largest number of relevant articles. Among all the journals in this field, PLoS ONE and Computers and Electronics in Agriculture ranked first and second in terms of annual publications and co-citation frequency. The present character, hot spots, and future research trends of AI technologies in food safety research were determined. Furthermore, based on our analyses, we provide researchers, practitioners, and policymakers with the big picture of research on AI in food safety across the whole process, from precision agriculture to precision nutrition, through 28 enlightening articles.

A Survey of Convolutional Neural Network in Breast Cancer

Problems

For people all over the world, cancer is one of the most feared diseases. Cancer is one of the major obstacles to improving life expectancy in countries around the world and one of the biggest causes of death before the age of 70 in 112 countries. Among all kinds of cancers, breast cancer is the most common cancer for women. The data showed that female breast cancer had become one of the most common cancers.

Aims

A large number of clinical trials have proved that if breast cancer is diagnosed at an early stage, it could give patients more treatment options and improve the treatment effect and survival ability. Based on this situation, there are many diagnostic methods for breast cancer, such as computer-aided diagnosis (CAD).

Methods

We complete a comprehensive review of the diagnosis of breast cancer based on the convolutional neural network (CNN) after reviewing a sea of recent papers. Firstly, we introduce several different imaging modalities. The structure of CNN is given in the second part. After that, we introduce some public breast cancer data sets. Then, we divide the diagnosis of breast cancer into three different tasks: 1. classification; 2. detection; 3. segmentation.

Conclusion

Although this diagnosis with CNN has achieved great success, there are still some limitations. (i) There are too few good data sets. A good public breast cancer dataset needs to involve many aspects, such as professional medical knowledge, privacy issues, financial issues, dataset size, and so on. (ii) When the data set is too large, the CNN-based model needs a sea of computation and time to complete the diagnosis. (iii) It is easy to cause overfitting when using small data sets.

Recent advances in shelf life prediction models for monitoring food quality

Each year, 1.3 billion tons of food is lost due to spoilage or loss in the supply chain, accounting for approximately one third of global food production. This requires a manufacturer to provide accurate information on the shelf life of the food in each stage. Various models for monitoring food quality have been developed and applied to predict food shelf life. This review classified shelf life models and detailed the application background and characteristics of commonly used models to better understand the different uses and aspects of the commonly used models. In particular, the structural framework, application mechanisms, and numerical relationships of commonly used models were elaborated. In addition, the study focused on the application of commonly used models in the food field. Besides predicting the freshness index and remaining shelf life of food, the study addressed aspects such as food classification (maturity and damage) and content prediction. Finally, further promotion of shelf life models in the food field, use of multivariate analysis methods, and development of new models were foreseen. More reliable transportation, processing, and packaging methods could be screened out based on real-time food quality monitoring.

Evaluation of vegetable sauerkraut quality during storage based on convolution neural network

Vegetable sauerkraut is a traditional fermented food. Due to oxidation reactions that occur during storage, the quality and flavor in different periods will change. In this study, the quality evaluation and flavor characteristics of 13 groups of vegetable sauerkraut samples with different storage time were analyzed by using physical and chemical parameters combined with electronic nose. Photographs of samples of various periods were collected, and a convolutional neural network (CNN) framework was established. The relationship between total phenol oxidative decomposition and flavor compounds was linearly negatively correlated. The vegetable sauerkraut during storage can be divided into three categories (full acceptance period, acceptance period and unacceptance period) by principal component analysis and Fisher discriminant analysis. The CNN parameters were fine-tuned based on the classification results, and its output results can reflect the quality changes and flavor characteristics of the samples, and have better fitting, prediction capabilities. After 50 epochs of the model, the accuracy of three sets of data namely training set, validation set and test set recorded 94%, 85% and 93%, respectively. In addition, the accuracy of CNN in identifying different quality sauerkraut was 95.30%. It is proved that the convolutional neural network has excellent performance in predicting the quality of Szechuan Sauerkraut with high reliability.

ResNet Model Automatically Extracts and Identifies FT-NIR Features for Geographical Traceability of Polygonatum kingianum

Medicinal plants have incredibly high economic value, and a practical evaluation of their quality is the key to promoting industry development. The deep learning model based on residual convolutional neural network (ResNet) has the advantage of automatic extraction and the recognition of Fourier transform near-infrared spectroscopy (FT-NIR) features. Models are difficult to understand and interpret because of unknown working mechanisms and decision-making processes. Therefore, in this study, artificial feature extraction methods combine traditional partial least squares discriminant analysis (PLS-DA) and support vector machine (SVM) models to understand and compare deep learning models. The results show that the ResNet model has significant advantages over traditional models in feature extraction and recognition. Secondly, preprocessing has a great impact on the feature extraction and feature extraction, and is beneficial for improving model performance. Competitive adaptive reweighted sampling (CARS) and variable importance in projection (VIP) methods screen out more feature variables after preprocessing, but the number of potential variables (LVs) and successive projections algorithm (SPA) methods obtained is fewer. The SPA method only extracts two variables after preprocessing, causing vital information to be lost. The VIP feature of traditional modelling yields the best results among the four methods. After spectral preprocessing, the recognition rates of the PLS-DA and SVM models are up to 90.16% and 88.52%. For the ResNet model, preprocessing is beneficial for extracting and identifying spectral image features. The ResNet model based on synchronous two-dimensional correlation spectra has a recognition accuracy of 100%. This research is beneficial to the application development of the ResNet model in foods, spices, and medicinal plants.

Neural network in food analytics

Neural network (i.e. deep learning, NN)-based data analysis techniques have been listed as a pivotal opportunity to protect the integrity and safety of the global food supply chain and forecast $11.2 billion in agriculture markets. As a general-purpose data analytic tool, NN has been applied in several areas of food science, such as food recognition, food supply chain security and omics analysis, and so on. Therefore, given the rapid emergence of NN applications in food safety, this review aims to provide a comprehensive overview of the NN application in food analysis for the first time, focusing on domain-specific applications in food analysis by introducing fundamental methodology, reviewing recent and notable progress, and discussing challenges and potential pitfalls. NN demonstrated that it has a bright future through effective collaboration between food specialist and the broader community in the food field, for example, superiority in food recognition, sensory evaluation, pattern recognition of spectroscopy and chromatography. However, major challenges impeded NN extension including void in the food scientist-friendly interface software package, incomprehensible model behavior, multi-source heterogeneous data, and so on. The breakthrough from other fields proved NN has the potential to offer a revolution in the immediate future.

ResNet and MaxEnt modeling for quality assessment of Wolfiporia cocos based on FT-NIR fingerprints

As a fungus with both medicinal and edible value, Wolfiporia cocos (F. A. Wolf) Ryvarden & Gilb. has drawn more public attention. Chemical components' content fluctuates in wild and cultivated W. cocos, whereas the accumulation ability of chemical components in different parts is different. In order to perform a quality assessment of W. cocos, we proposed a comprehensive method which was mainly realized by Fourier transform near-infrared (FT-NIR) spectroscopy and ultra-fast liquid chromatography (UFLC). A qualitative analysis means was built a residual convolutional neural network (ResNet) to recognize synchronous two-dimensional correlation spectroscopy (2DCOS) images. It can rapidly identify samples from wild and cultivated W. cocos in different parts. As a quantitative analysis method, UFLC was used to determine the contents of three triterpene acids in 547 samples. The results showed that a simultaneous qualitative and quantitative strategy could accurately evaluate the quality of W. cocos. The accuracy of ResNet models combined synchronous FT-NIR 2DCOS in identifying wild and cultivated W. cocos in different parts was as high as 100%. The contents of three triterpene acids in Poriae Cutis were higher than that in Poria, and the one with wild Poriae Cutis was the highest. In addition, the suitable habitat plays a crucial role in the quality of W. cocos. The maximum entropy (MaxEnt) model is a common method to predict the suitable habitat area for W. cocos under the current climate. Through the results, we found that suitable habitats were mostly situated in Yunnan Province of China, which accounted for approximately 49% of the total suitable habitat area of China. The research results not only pave the way for the rational planting in Yunnan Province of China and resource utilization of W. cocos, but also provide a basis for quality assessment of medicinal fungi.

Artificial Intelligence Technology Driven Environmental Factors Extraction and Analysis Method in Traditional Clothing Handicraft

The application of artificial intelligence (AI) technology in the field of clothes can provide a good development mode and system under the social context of AI technology development. AI provides help for the development of intelligent clothing. Intelligent clothing is a high-tech product that integrates intelligent technology and clothing. It combines cutting-edge technologies in electronic information technology, sensor technology, textile science, and material science. In the extraction and analysis of environmental factors in clothing handicraft, AI technology has a considerable application prospect and a certain development potential. In order to improve the accuracy of environmental factors extraction in clothing handicraft, this paper uses convolutional neural network (CNN) to extract and analyze environmental factors in traditional clothing handicraft. We carried out experiments on the extraction of environmental factors in clothing handicrafts with pure color, few patterns, patterns, and complex background. The experimental results show that the CNN has a good effect on the extraction of environmental factors in clothing handicraft under different backgrounds. In addition, the model in this paper has good stability, accuracy, and feature extraction speed, which has high practical value and research significance.

Applications of machine learning techniques for enhancing nondestructive food quality and safety detection

In considering the need of people all over the world for high-quality food, there has been a recent increase in interest in the role of nondestructive and rapid detection technologies in the food industry. Moreover, the analysis of data acquired by most nondestructive technologies is complex, time-consuming, and requires highly skilled operators. Meanwhile, the general applicability of various chemometric or statistical methods is affected by noise, sample, variability, and data complexity that vary under various testing conditions. Nowadays, machine learning (ML) techniques have a wide range of applications in the food industry, especially in nondestructive technology and equipment intelligence, due to their powerful ability in handling irrelevant information, extracting feature variables, and building calibration models. The review provides an introduction and comparison of machine learning techniques, and summarizes these algorithms as traditional machine learning (TML), and deep learning (DL). Moreover, several novel nondestructive technologies, namely acoustic analysis, machine vision (MV), electronic nose (E-nose), and spectral imaging, combined with different advanced ML techniques and their applications in food quality assessment such as variety identification and classification, safety inspection and processing control, are presented. In addition to this, the existing challenges and prospects are discussed. The result of this review indicates that nondestructive testing technologies combined with state-of-the-art machine learning techniques show great potential for monitoring the quality and safety of food products and different machine learning algorithms have their characteristics and applicability scenarios. Due to the nature of feature learning, DL is one of the most promising and powerful techniques for real-time applications, which needs further research for full and wide applications in the food industry.

Using Convolutional Neural Networks to Derive Neighborhood Built Environments from Google Street View Images and Examine Their Associations with Health Outcomes

Built environment neighborhood characteristics are difficult to measure and assess on a large scale. Consequently, there is a lack of sufficient data that can help us investigate neighborhood characteristics as structural determinants of health on a national level. The objective of this study is to utilize publicly available Google Street View images as a data source for characterizing built environments and to examine the influence of built environments on chronic diseases and health behaviors in the United States. Data were collected by processing 164 million Google Street View images from November 2019 across the United States. Convolutional Neural Networks, a class of multi-layer deep neural networks, were used to extract features of the built environment. Validation analyses found accuracies of 82% or higher across neighborhood characteristics. In regression analyses controlling for census tract sociodemographics, we find that single-lane roads (an indicator of lower urban development) were linked with chronic conditions and worse mental health. Walkability and urbanicity indicators such as crosswalks, sidewalks, and two or more cars were associated with better health, including reduction in depression, obesity, high blood pressure, and high cholesterol. Street signs and streetlights were also found to be associated with decreased chronic conditions. Chain link fence (physical disorder indicator) was generally associated with poorer mental health. Living in neighborhoods with a built environment that supports social interaction and physical activity can lead to positive health outcomes. Computer vision models using manually annotated Google Street View images as a training dataset were able to accurately identify neighborhood built environment characteristics. These methods increases the feasibility, scale, and efficiency of neighborhood studies on health.

Application of Intelligent Taste Analysis Based on Random Forest Algorithm in Food Quality Inspection

Food safety is a major concern that has an impact on the national economy and people's lives. The food industry has grown in quality and innovation in tandem with the rapid development of the economy and society. The emergence of new food technologies, as well as changes in dietary habits, has increased public concern about food safety. With the emergence of various counterfeit and substandard products, food quality and safety testing have become even more important. Traditional testing methods rely on sensory analysis and physical and chemical analysis. This approach is subjective and poorly adapted to the general public. It requires a high level of technical operation and is difficult to carry out on a large scale. To address this situation, this paper proposes an intelligent approach to food safety quality testing. The core idea is, first, to use sensors to collect data on the various components of the sample to be tested. Second, the random forest (RF) model used in this paper is trained. Third, the trained model is used to classify and identify the test samples. Based on the classification results, a conclusion is drawn as to whether the food product is a variant or a counterfeit. The advantage of this study is that the training model used is a weighted RF algorithm based on mutual information. The correlation between any two decision trees is calculated using mutual information, and for the more correlated decision trees, only the one with the highest evaluation accuracy is retained to form a new RF, and the evaluation accuracy is converted into voting weights, resulting in an RF model with less redundancy and higher evaluation accuracy. The experimental results show that the method used in this paper can successfully identify spoiled or counterfeit products and has good practicality.

Identification and quantification of adulterated honey by Raman spectroscopy combined with convolutional neural network and chemometrics

In this study, Raman spectroscopy combined with convolutional neural network (CNN) and chemometrics was used to achieve the identification and quantification of honey samples adulterated with high fructose corn syrup, rice syrup, maltose syrup and blended syrup, respectively. The shallow CNNs utilized to analyze honey mixed with single-variety syrup classified samples into four categories by the adulteration concentration with more than 97% accuracy, and the general CNN model for simultaneously detecting honey adulterated with any type of syrup obtained an accuracy of 94.79%. The established CNNs had the best performance compared with several chemometric classification algorithms. In addition, partial least square regression (PLS) successfully predicted the purity of honey mixed with single syrup, while coefficients of determination and root mean square errors of prediction were greater than 0.98 and less than 3.50, respectively. Therefore, the proposed methods based on Raman spectra have important practical significance for food safety and quality control of honey products.

Applications of knowledge graphs for food science and industry

The deployment of various networks (e.g., Internet of Things [IoT] and mobile networks), databases (e.g., nutrition tables and food compositional databases), and social media (e.g., Instagram and Twitter) generates huge amounts of food data, which present researchers with an unprecedented opportunity to study various problems and applications in food science and industry via data-driven computational methods. However, these multi-source heterogeneous food data appear as information silos, leading to difficulty in fully exploiting these food data. The knowledge graph provides a unified and standardized conceptual terminology in a structured form, and thus can effectively organize these food data to benefit various applications. In this review, we provide a brief introduction to knowledge graphs and the evolution of food knowledge organization mainly from food ontology to food knowledge graphs. We then summarize seven representative applications of food knowledge graphs, such as new recipe development, diet-disease correlation discovery, and personalized dietary recommendation. We also discuss future directions in this field, such as multimodal food knowledge graph construction and food knowledge graphs for human health.

Imaging Spectroscopy and Machine Learning for Intelligent Determination of Potato and Sweet Potato Quality

Imaging spectroscopy has emerged as a reliable analytical method for effectively characterizing and quantifying quality attributes of agricultural products. By providing spectral information relevant to food quality properties, imaging spectroscopy has been demonstrated to be a potential method for rapid and non-destructive classification, authentication, and prediction of quality parameters of various categories of tubers, including potato and sweet potato. The imaging technique has demonstrated great capacities for gaining rapid information about tuber physical properties (such as texture, water binding capacity, and specific gravity), chemical components (such as protein, starch, and total anthocyanin), varietal authentication, and defect aspects. This paper emphasizes how recent developments in spectral imaging with machine learning have enhanced overall capabilities to evaluate tubers. The machine learning algorithms coupled with feature variable identification approaches have obtained acceptable results. This review briefly introduces imaging spectroscopy and machine learning, then provides examples and discussions of these techniques in tuber quality determinations, and presents the challenges and future prospects of the technology. This review will be of great significance to the study of tubers using spectral imaging technology.