Evaluation of visible contamination on power grid insulators using convolutional neural networks

Electrical Engineering, Год журнала: 2023, Номер 105(6), С. 3881 - 3894

Опубликована: Июль 14, 2023

Язык: Английский

---

Detection of fish freshness using artificial intelligence methods

European Food Research and Technology, Год журнала: 2023, Номер 249(8), С. 1979 - 1990

Опубликована: Апрель 27, 2023

Язык: Английский

---

Wildfire fuels mapping through artificial intelligence-based methods: A review

Earth-Science Reviews, Год журнала: 2025, Номер unknown, С. 105064 - 105064

Опубликована: Фев. 1, 2025

Язык: Английский

---

Beef Quality Classification with Reduced E-Nose Data Features According to Beef Cut Types

Sensors, Год журнала: 2023, Номер 23(4), С. 2222 - 2222

Опубликована: Фев. 16, 2023

Ensuring safe food supplies has recently become a serious problem all over the world. Controlling quality, spoilage, and standing time for products with short shelf life is quite difficult problem. However, electronic noses can make these controls possible. In this study, which aims to develop different approach solution of problem, nose data obtained from 12 beef cuts were classified. dataset, there are four classes (1: excellent, 2: good, 3: acceptable, 4: spoiled) indicating quality. The classifications performed separately each cut shapes. ANOVA method was used determine active features in dataset features. same classification processes carried out by using three selected method. Three machine learning methods, Artificial Neural Network, K Nearest Neighbor, Logistic Regression, frequently literature, classifications. experimental studies, accuracy 100% as result ANN combining tables dataset.

Язык: Английский

---

Segmentation of dry bean (Phaseolus vulgaris L.) leaf disease images with U-Net and classification using deep learning algorithms

European Food Research and Technology, Год журнала: 2023, Номер 249(10), С. 2543 - 2558

Опубликована: Июнь 26, 2023

Язык: Английский

---

Fisheye freshness detection using common deep learning algorithms and machine learning methods with a developed mobile application

European Food Research and Technology, Год журнала: 2024, Номер 250(7), С. 1919 - 1932

Опубликована: Апрель 18, 2024

Abstract Fish is commonly ingested as a source of protein and essential nutrients for humans. To fully benefit from the proteins substances in fish it crucial to ensure its freshness. If stored an extended period, freshness deteriorates. Determining can be done by examining eyes, smell, skin, gills. In this study, artificial intelligence techniques are employed assess The author’s objective evaluate analyzing eye characteristics. achieve this, we have developed combination deep machine learning models that accurately classify fish. Furthermore, application utilizes both learning, instantly detect any given sample was created. Two algorithms (SqueezeNet, VGG19) were implemented extract features image data. Additionally, five levels samples applied. Machine include (k-NN, RF, SVM, LR, ANN). Based on results, inferred employing VGG19 model feature selection conjunction with Artificial Neural Network (ANN) classification yields most favorable success rate 77.3% FFE dataset. Graphical

Язык: Английский

---

An efficient deepfake video detection using robust deep learning

Heliyon, Год журнала: 2024, Номер 10(5), С. e25757 - e25757

Опубликована: Фев. 9, 2024

The creation and manipulation of synthetic images have evolved rapidly, causing serious concerns about their effects on society. Although there been various attempts to identify deep fake videos, these approaches are not universal. Identifying misleading deepfakes is the first step in preventing them from spreading social media sites. We introduce a unique deep-learning technique fraudulent clips. Most deepfake identifiers currently focus identifying face exchange, lip synchronous, expression modification, puppeteers, other factors. However, exploring consistent basis for all forms videos real-time forensics challenging. propose hybrid that takes input successive targeted frames, then feeds frames ResNet-Swish-BiLSTM, an optimized convolutional BiLSTM-based residual network training classification. This proposed method helps artifacts do seem real. To assess robustness our model, we used open detection challenge dataset (DFDC) Face Forensics collections (FF++). achieved 96.23% accuracy when using FF++ digital record. In contrast, attained 78.33% aggregated records DFDC. performed extensive experiments believe provides more significant results than existing techniques.

Язык: Английский

---

Benchmarking analysis of CNN models for bread wheat varieties

European Food Research and Technology, Год журнала: 2022, Номер 249(3), С. 749 - 758

Опубликована: Ноя. 16, 2022

Язык: Английский

---

Digital Technology Increases the Sustainability of Cross-Border Agro-Food Supply Chains: A Review

Agriculture, Год журнала: 2024, Номер 14(6), С. 900 - 900

Опубликована: Июнь 6, 2024

The increasing prominence of climate change, geopolitical crises, and global economic slowdown highlights the challenges structural deficiencies traditional cross-border agro-food supply chains. As a result, there has been growing consensus on need to leverage digital technology rebuild innovate safe, stable, sustainable food system. This study assessed knowledge progress development trends in chains enabled by technology. A total 352 authoritative papers from core Web Science database were selected for analysis. Citespace tool was utilized visually examine research elements. findings reveal that outcomes this territory experienced significant period rapid growth, particularly after 2020. Sustainability IEEE Access are journals with highest second-highest number publications. China France National Institute countries institutions largest publications field. hotspots mainly application technologies, safety, chain system model innovation. In past ten years, gone through three stages: precise timeliness orientation, intelligent strategic decision-making predictability orientation. We further construct ‘antecedent–practice–performance’ conceptual framework sustainability technology-enabled chain. Finally, paper presents potential directions territory, focusing four aspects: method, mechanism, topic, frontier.

Язык: Английский

---

Deep learning and evolutionary intelligence with fusion-based feature extraction for classification of wheat varieties

European Food Research and Technology, Год журнала: 2025, Номер unknown

Опубликована: Апрель 30, 2025

Язык: Английский

---