作者：
Zirong LvSchool of Automation
China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China
Crop disease diagnosis is of great significance for crop yield and agricultural production. Deep learning methods have become a major research direction for solving crop disease diagnosis. However, recent studies have...
Crop disease diagnosis is of great significance for crop yield and agricultural production. Deep learning methods have become a major research direction for solving crop disease diagnosis. However, recent studies have shown that deploying complex neural network models on mobile devices poses a significant challenge to mobile hardware performance. To address this issue, this paper proposes a lightweight network model based on the Transformer and deploys it on mobile devices. Compared to other lightweight network models, our approach significantly reduces computational complexity using Transformer modules and exhibits a particular advantage in recognition accuracy. It achieves a recognition rate of 97.76% for ten categories of tomato diseases, while the model size is only 3.86 MB. Furthermore, we deploy the model on embedded devices and apply it to tomato crop recognition, enabling farmers to monitor crop conditions in real-time and realizing smart agriculture.
Recent advancements in Novel View Synthesis using neural radiance fields and 3D Gaussian splatting have demon-strated promising results. However, these methods conventionally rely on accurate initial poses and point c...
详细信息
ISBN:
(数字)9798350368604
ISBN:
(纸本)9798350368611
Recent advancements in Novel View Synthesis using neural radiance fields and 3D Gaussian splatting have demon-strated promising results. However, these methods conventionally rely on accurate initial poses and point clouds from the model of structure from motion, which poses challenges in sparse input views. To addressing these drawbacks, we introduces an novel method for view synthesis based on 3D Gaussian splatting. Our approach operates on sparse input views without the prerequisite of pose priors, achieving reconstruction and real-time rendering of novel perspectives. for the deficiency of camera poses and point cloud, we employ an end- to-end deep learning model to recover the 3D structure from a few of 2D images to initialize the 3D Gaussian point cloud. In addressing the issue with few-shot, we utilize the differentiable rasterization module to render depth map and apply regulation on it. Additionally, we incorporate supervision between input views by interpolating to generate novel perspectives during training. Experimental results demonstrate that our proposed method significantly enhances the capabilities of 3D Gaussian splatting for novel view synthesis in sparse views and without the necessity of camera pose priors.
The issue of H∞ state estimation for neural networks with time-varying delays is investigated in this study. Firstly, an augmented Lyapunov-Krasovskii functional (LKF) with two delay-product-type terms is constructed...
详细信息
In response to the national dual carbon policy, carbon capture, utilization and storage (CCUS) are currently attracting much attention. This paper proposes optimal CCUS planning for multi-energy system. In this multi-...
详细信息
Dear Editor,This letter is concerned with dealing with the great discrepancy between near-infrared(NIR)and visible(VS)image fusion via color distribution preserved generative adversarial network(CDP-GAN).Different fro...
详细信息
Dear Editor,This letter is concerned with dealing with the great discrepancy between near-infrared(NIR)and visible(VS)image fusion via color distribution preserved generative adversarial network(CDP-GAN).Different from the global discriminator in prior GAN,conflict of preserving NIR details and VS color is resolved by introducing an attention guidance mechanism into the ***。
Affected by the global New Crown Pneumonia epidemic, energy prices in the global market continue to rise. The high quality production and energy saving of steel companies have become an important task. The blast furna...
详细信息
Affected by the global New Crown Pneumonia epidemic, energy prices in the global market continue to rise. The high quality production and energy saving of steel companies have become an important task. The blast furnace is the front-end core of the steel manufacturing process. The blast furnace gas utilization rate can effectively characterize the internal airflow distribution, blast furnace operation and energy consumption level of the blast furnace. The blast furnace Gas Utilization Rate (GUR) can effectively characterize the internal airflow distribution, blast furnace operation and energy consumption level. Most of the existing studies are based on data-driven models. The shallow neural network model is selected. But blast furnace iron making has complex uncertainty. Massive data samples under industrial information technology need to be processed. The robustness and generalization ability of the prediction model are not satisfactory. To address the above problems, this paper proposes a prediction model based on NGO-LSTM regression. The model parameter searchs can be intelligently. It achieves high-precision prediction results for massive data samples. Firstly, the selection of feature parameters is completed by maximal information coefficient. Secondly, the strong coupling of blast furnace ground needs to fully predict the relationship between the characteristic parameters. The Long Short-Term Memory (LSTM) neural network with certain memory capability is selected. And multiple parameters of this neural network model are optimized by the Northern Goshawk Optimization (NGO) algorithm. An NGO-LSTM regression prediction model is established. In this paper, experiments are carried out using actual production data. The experimental results show that the proposed method can accurately predict the blast furnace gas utilization rate. This can provide a reference for improving blast furnace product quality, reducing costs and increasing efficiency.
Facial expression recognition plays a key role in promoting the development of comprehensive intelligence and building friendly human-computer interaction. Due to the interference of feature noise in expression data, ...
详细信息
In recent years, the semantic segmentation of 3D point cloud has received increasing attention the field of computer vision, because 3D point cloud can better reflect our 3D space. Because of the unstructured and diso...
详细信息
In the cement production process, the decomposer outlet temperature control faces challenges such as large time delay and multiple disturbances. A general control system cannot handle time delay and disturbances effec...
详细信息
ISBN:
(数字)9798350368604
ISBN:
(纸本)9798350368611
In the cement production process, the decomposer outlet temperature control faces challenges such as large time delay and multiple disturbances. A general control system cannot handle time delay and disturbances effectively at the same time. Therefore, in this paper, we consider the problem of disturbances in time-delay systems and suppress the disturbances present in the outlet temperature control of the decomposer by combining equivalent-input-disturbance (EID) and Model Predictive control (MPC), where EID can estimate the disturbances and apply them to the input channels of the system, which do not require any information about the disturbances. Finally simulations are carried out in the catabolizer system and the results indicate that this system is able to suppress the perturbations effectively in time delay systems.
This paper describes a parameter of voltage sensitivity to recognize the performance differences of tag antennas for inductively coupled RFID systems. Based on the equivalent circuit model of the RFID tag and reader, ...
详细信息
暂无评论