VCPC:虚拟对比约束和原型校准为少数射击类-增量植物疾病分类
Lunhong Lou1,2,3, Jianwu Lin2,3, Lin You1,2,3
1College of Big Data and Information Engineering, Guizhou University, Guiyang, 550025, China.
Plant methods
|July 31, 2025
概括
这项研究引入了VCPV,这是一种实时识别植物疾病的新方法,可以适应具有有限数据的新疾病类型. 它通过能够可持续,准确地识别新型植物病来加强作物监测.
科学领域:
- 农业科学 农业科学
- 计算机科学 计算机科学
- 机器学习 机器学习
背景情况:
- 深度学习推进了植物疾病识别,但与实时新型疾病检测扎.
- 现有的系统大多处于离线状态,并且在几次射击条件下缺乏适应新类疾病的适应性.
- 实时作物监测需要逐步适应的模型,以进行少数拍摄的阶级增量学习 (FSCIL).
研究的目的:
- 在FSCIL条件下引入VCPV (具有原型载体校准的虚拟对比约束) 以进行可持续的植物疾病分类.
- 开发一种方法,使实时作物监测系统能够识别具有有限例子的新植物疾病.
- 提高植物疾病识别模型在动态环境中的适应性和准确性.
主要方法:
- 该VCPV方法包括两个阶段:基础班级培训和增量培训.
- 虚拟对比类约束 (VCC) 模块增强了基础类的学习,并为新类分配了嵌入空间.
- 一个原型校准嵌入 (PCE) 模块将新类别与现有类别区分开来,优化原型空间和识别精度.
主要成果:
- 在5向5射和3向5射设置中,VCPV在PlantVillage数据集上实现了最先进的准确性.
- 该方法在CIFAR-100数据集上显示出有希望的性能.
- 视觉化结果证实了该战略在细粒度,可持续的疾病识别方面的有效性.
结论:
- VCPV提供了一种可持续的解决方案,用于植物疾病分类,在少数阶段阶级增量学习场景中进行分类.
- 该方法显著提高了人工智能系统实时监测新型疾病作物的能力.
- 在植物疾病监测和农业应用中,VCPV具有推动FSCIL发展的潜力.
相关概念视频
02:16Light Acquisition
8.6K
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
8.6K
01:59Plant Breeding and Biotechnology
19.8K
Crop cultivation has a long history in human civilization, with records showing the cultivation of cereal plants beginning at around 8000 BC. This early plant breeding was developed primarily to provide a steady supply of food.
19.8K
01:31Classification of Systems-II
242
Continuous-time systems have continuous input and output signals, with time measured continuously. These systems are generally defined by differential or algebraic equations. For instance, in an RC circuit, the relationship between input and output voltage is expressed through a differential equation derived from Ohm's law and the capacitor relation,
242
01:26Classification of Systems-I
312
Linearity is a system property characterized by a direct input-output relationship, combining homogeneity and additivity.
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
Homogeneity dictates that if an input x(t) is multiplied by a constant c, the output y(t) is multiplied by the same constant. Mathematically, this is expressed as:
312
01:17Classification of Illness
8.0K
The meaning of illness is individualized to each person who experiences an alteration in health. In contrast, disease is a medical term indicating a pathological change in the structure and function of the body or mind. It is a condition that has specific symptoms and boundaries.
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
8.0K
01:20Calibration Curves: Linear Least Squares
2.3K
A calibration curve is a plot of the instrument's response against a series of known concentrations of a substance. This curve is used to set the instrument response levels, using the substance and its concentrations as standards. Alternatively, or additionally, an equation is fitted to the calibration curve plot and subsequently used to calculate the unknown concentrations of other samples reliably.
For data that follow a straight line, the standard method for fitting is the linear...
For data that follow a straight line, the standard method for fitting is the linear...
2.3K