Landscape gardening design is an important part of urban ecology and urban image, which is of great significance for urban plant protection. Plant diseases are a major challenge that must be faced in the conservation of landscape plants. Having timely and comprehensive information about the types and developmental stages of plant diseases is crucial for effective disease control. Traditional plant disease differentiation, which relies on manual inspection by professionals, is both time-consuming and labor-intensive. The rapid development of deep artificial neural network technology, especially the rise of computer vision technology, provides efficient and accurate technical means for plant disease identification. In this paper, we introduce the Interactive Bilinear Transformer Network (IBTN) model, which utilizes fine-grained recognition technology for landscape plant disease identification. We validate the effectiveness of model using several public datasets, demonstrating its competitive performance.

中文翻译：

---

基于双线性变压器交互式神经网络的园艺设计植物病害细粒度识别和保护方法

景观园林设计是城市生态和城市形象的重要组成部分，对于城市植物保护具有重要意义。植物病害是园林植物保护必须面对的重大挑战。及时、全面地了解植物病害的类型和发育阶段对于有效控制病害至关重要。传统的植物病害鉴别依靠专业人员人工检查，既费时又费力。深度人工神经网络技术的快速发展，特别是计算机视觉技术的兴起，为植物病害识别提供了高效、准确的技术手段。在本文中，我们介绍了交互式双线性变压器网络（IBTN）模型，该模型利用细粒度识别技术进行景观植物病害识别。我们使用多个公共数据集验证模型的有效性，展示其竞争性能。