The integration of machine learning technologies into print quality control systems represents a significant advancement in modern printing production. This article examines the application of artificial intelligence methods and computer vision algorithms for automated defect detection, colour consistency monitoring, and real-time quality assessment in printing processes.
This paper considers the problem of adaptive optimization of the training dataset for neural networks in image classification tasks. It has been shown that using the full training dataset is not the best solution. Different training samples have different value for the model. Some samples are representative, some are important for class boundaries, some preserve diversity, and some may be noisy or suspicious. A method for adaptive formation of the active training subset has been proposed. Unlike traditional methods, the proposed approach evaluates the role of each sample during training.
This review paper provides a comprehensive comparative analysis of abandoned object detection algorithms, specifically contrasting rule-based approaches with artificial intelligence (AI-based) models. The manuscript synthesizes existing research, technical documentation, and publicly available benchmark data to evaluate the applicability of these approaches in dynamic video surveillance environments.
This paper presents a method for optimizing deep learning models used in real-time military equipment recognition systems. A key contribution of the work is the application of exponential data augmentation, which significantly increases dataset diversity without requiring additional real-world data. The augmented dataset has been used to train a YOLO-based object detection model capable of recognizing various types of military vehicles, including tanks, armored personnel carriers, and infantry fighting vehicles.
The article addresses the urgent scientific and practical problem of overcoming the "cold start" effect in the development and deployment of Natural Language Processing (NLP) systems aimed at monitoring public opinion and sentiment analysis of Ukrainian-language content. The critical shortage of representative, balanced, and pre-labeled datasets that accurately reflect the specifics of social, economic, and political processes in modern Ukrainian society is identified as the main barrier to integrating advanced neural network solutions.
Summary. Automation of database indexing is a crucial component of modern database management systems that enhances search performance, scalability, and relevance in large-scale data environments. This paper explores the application of deep learning algorithms for building and optimizing vector indexes capable of automatic adaptation to changes in data structure and query patterns. An experimental comparison was conducted between traditional indexing methods (B-Tree, GIN in PostgreSQL) and vector-based indexing using Sentence-BERT embeddings implemented in FAISS and Milvus systems.
Detecting plant diseases on time is imperative to improve agricultural productivity and mitigate economic losses. This study presents a novel artificial intelligence framework for classifying plant diseases, combining advanced deep learning technologies with a hyperparameter optimization strategy. Specifically, we employed a hybrid architecture that concatenates two pre-trained models, MobileNetV2 and DenseNet201, enriched with custom layers developed by the researchers. Bayesian Optimization was employed to enhance model performance, focusing on four critical hyperp
Emerging as a crucial tool for companies in many industries, emotion recognition offers a greater understanding of consumer views of goods and services, hence strengthening client connections, improving service delivery, and guiding emotionally intelligent marketing strategies. Particularly with the use of transfer learning techniques, end-to-end image-based emotion categorization has gained popularity in the last several years. In tackling such challenging tasks, deep learning models have shown great efficacy. Though reading emotions from visual data presents diffic
The goal of the article is to provide a methodology of improving images quality in low-light conditions trough fusion of data received from telecamera and thermal camera. Data from thermal camera uses for compensation of significant illumination reduction in poor lighting conditions and allow keep required level of information. Proposed method establishes dynamic regulation of fusion coefficients depending on brightness level to minimize artifacts, increase edge sharpness, and improve object detectability.
Visual object tracking has recently benefited from the adoption of transformer architectures, which provide strong modeling capacity but incur high computational and memory costs, limiting real-time deployment. Existing efficiency-focused trackers primarily address this challenge through architectural redesign, often trading accuracy for speed.