ALAndrii Lutsiuk

Andrii Lutsiuk

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Andrii Lutsiuk has a total of 10 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Prospects for the application of the YOLOv11 model for automatic control of combine harvesters based on the detection of damaged grains

    by A. Ya. Kuzmych, M. M. Aneliak, S. P. Stepanenko, A. V. Lutsiuk

    Purpose. Improving the efficiency of grain harvesting by developing scientific and technical principles for an automatic combine harvester control system based on the YOLOv11 model, which minimizes grain damage losses through high-precision real-time detection. Methods. Methods of data analysis, collection and preparation, digital image segmentation, deep learning of convolutional neural networks were used. Software for digital image analysis was developed in the Python programming language using the computer vision libraries OpenCV and PyTorch. Results. The concept and algorithm of the automation system for the control of a combine harvester were developed. The influence of the parameters of the working bodies of the combine harvester on grain injury was studied. A dataset was collected and the YOLOv11 model was trained for the classification and segmentation of winter wheat grain material. The trained model was tested by detecting objects and segmenting them on images of real samples of grain materials. Conclusions. The research confirms that the implementation of an automatic control system for combine harvesters based on object detection is not only technically feasible, but also economically viable. The problem of grain damage, which leads to significant financial losses, can be solved using technologies capable of providing constant and accurate quality control in real time. The YOLOv11 model, due to its speed and architectural innovations, meets the requirements necessary for dynamic control of a combine harvester. In addition, its architecture, specially designed for improved detection of small and partially hidden objects, solves key challenges in visual analysis of grain material flows. The conceptual model of the system, based on peripheral computing, can operate autonomously in field conditions, making it independent of the network infrastructure. Testing of the model demonstrated a high level of detection of grain mixture objects. According to the results of testing on the Jetson AGX Orin GPU, the processing time for a 1280x1280 image was 90.6 ms. Keywords: winter wheat, combine harvester, convolutional neural network, machine vision.

  • Open Access

    Substantiating the YOLO11 architecture for determining the fractional composition of winter wheat grain mixtures

    by Serhii Stepanenko, Alvian Kuzmych, Andrii Borys, Viktor Dnes, Serhii Kharchenko, Ivan Rogovskii, Gennadii Golub, Mykola Berezovyi, Andrii Lutsiuk

    This study’s object is the process of determining the fractional composition of winter wheat grain mixtures using computer vision and deep learning methods. The basic task that needs to be solved is the high complexity, subjectivity, and speed of determining the fractional composition of grain by conventional methods. The results obtained demonstrate the successful training and comparative analysis of several YOLO11-seg instance segmentation models on a specialized dataset deployed on the NVIDIA Jetson Orin platform. In particular, the YOLO11m-seg model with an image size of 640 × 640 pixels achieved the optimal compromise between accuracy and speed, achieving a Mask mAP50-95 index of 0.558 at an output speed of 62.5 ms/image. Training the YOLO11n-seg 1280 × 1280 model provided the best average segmentation accuracy (Mask mAP50-95 0.640) by increasing the performance of identifying objects of "complex" classes, which are key for accurate determination of the fractional composition. The results have made it possible to solve the problem under consideration through the empirically justified choice of architecture. Unlike hypothetical approaches, the study provides specific quantitative data on the performance of different YOLO11-seg architectures. That allowed for a reasonable selection of the model that best meets the requirements for accuracy and speed for practical deployment, solving the problem of uncertainty in the choice of architecture. The findings create the basis for automating grain quality control, increasing its efficiency, objectivity, and also significantly reducing labor intensity by automating routine operations. For practical use of the system, it is necessary to ensure stable lighting conditions, as well as the presence of a digital camera, a computer, and appropriate software


  • EFFICIENCY OF LLM INSTRUCTION FORMATS FOR CLASS IMBALANCE PROBLEMS IN TRAINING DATA FOR PREDICTIVE MONITORING SYSTEMS

    by Lviv Polytechnic National University, Andrii Lutsiuk

    The article examines approaches to formatting tabular data (HTML, XML, Markdown, CSV) for the subsequent generation of synthetic samples using large language models (LLM) in predictive monitoring tasks. Since real-world data are often characterized by class imbalance, generating additional samples helps improve training datasets, thereby enhancing the effectiveness of models. At the same time, an important issue arises regarding processing speed and query costs, which largely depend on how many input tokens are required by the chosen format for tabular data representation. The study analyzes computational resource consumption and query processing time for LLMs depending on the tabular data format. Although, according to research [1], HTML provides the highest level of accuracy, it also requires a significantly larger number of tokens due to its table representation format. This characteristic considerably increases the volume of input data and the overall query processing time. In contrast, less bulky formats (Markdown and CSV) require significantly fewer tokens, speeding up processing and reducing the cost of interaction with the model. A slight reduction in accuracy compared to HTML may be an acceptable trade-off, especially when the goal is to significantly expand the training dataset to compensate for the lack of examples of non-standard conditions. This approach proves to be effective in predictive monitoring systems, where response time and the volume of processed data directly affect the speed of anomaly detection and overall system resilience. The study results confirm that Markdown and CSV, due to their smaller input data volume, help reduce query processing time and the costs associated with generating synthetic training samples. At the same time, HTML and XML remain potentially useful in tasks where preserving complex structures and additional metadata is of utmost importance, although these formats require significantly more resources. Thus, the choice of a tabular data representation format should take into account the specific system requirements and operational environment characteristics, ranging from hardware limitations and token-based pricing to the required query processing time.