Artificial Intelligent Recognition for Multiple Supernumerary Teeth in Periapical Radiographs Based on Faster R-CNN and YOLOv8

Journal of Stomatology Oral and Maxillofacial Surgery, Год журнала: 2025, Номер unknown, С. 102293 - 102293

Опубликована: Фев. 1, 2025

Язык: Английский

YOLOv8 forestry pest recognition based on improved re-parametric convolution

Frontiers in Plant Science, Год журнала: 2025, Номер 16

Опубликована: Март 11, 2025

Introduction The ecological and economic impacts of forest pests have intensified, particularly in remote areas. Traditional pest detection methods are often inefficient inaccurate complex environments, posing significant challenges for effective management. Enhancing the efficiency accuracy under resource-limited conditions has thus become a critical issue. This study aims to address these by proposing an improved lightweight forestry algorithm, RSD-YOLOv8, based on YOLOv8. Methods To improve performance detection, we introduced several modifications YOLOv8 architecture. First, proposed RepLightConv replace conventional convolution HGNetV2, forming Rep-HGNetV2 backbone, which significantly reduces number model parameters. Additionally, neck was enhanced integrating slim-neck structure adding Dyhead module before output layer. Further optimization achieved through pruning, contributed additional lightweighting model. These improvements were designed balance with computational efficiency, deployment resource-constrained environments. Results experimental results demonstrate effectiveness RSD-YOLOv8 [email protected]:0.95(%) 88.6%, representing 4.2% improvement over original Furthermore, parameters reduced approximately 36%, operations decreased size 33%. indicate that not only enhances but also burden resource consumption. Discussion technology architectural this proven enhancing while minimizing requirements. model's ability operate efficiently areas limited resources makes it highly practical real-world applications. advancement holds positive implications agroforestry ecology supports broader goals intelligent sustainable development. Future work could explore further techniques application other domains requiring accurate systems.

Язык: Английский

DigiAgriApp: a client-server application to monitor field activities

Environmental Modelling & Software, Год журнала: 2025, Номер unknown, С. 106528 - 106528

Опубликована: Май 1, 2025

Язык: Английский