检索结果-内蒙古大学图书馆

Lightweight Object Detection Networks for UAV Aerial Images Based on YOLO

Chinese Journal of Electronics 2024年第4期33卷 997-1009页

作者： Yanshan LI Jiarong WANG Kunhua ZHANG Jiawei YI Miaomiao WEI Lirong ZHENG Weixin XIE ATR National Key Laboratory of Defense Technology Shenzhen University Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University

Existing high-precision object detection algorithms for UAV(unmanned aerial vehicle) aerial images often have a large number of parameters and heavy weight, which makes it difficult to be applied to mobile *** propose three YOLO-based lightweight object detection networks for UAVs, named YOLO-L, YOLO-S, and YOLO-M, respectively. In YOLO-L, we adopt a deconvolution approach to explore suitable upsampling rules during training to improve the detection accuracy. The convolution-batch normalization-Si LU activation function(CBS)structure is replaced with Ghost CBS to reduce the number of parameters and weight, meanwhile Maxpool maximum pooling operation is proposed to replace the CBS structure to avoid generating parameters and weight. YOLO-S greatly reduces the weight of the network by directly introducing CSPGhost Neck residual structures, so that the parameters and weight are respectively decreased by about 15% at the expense of 2.4% mAP. And YOLO-M adopts the CSPGhost Neck residual structure and deconvolution to reduce parameters by 5.6% and weight by 5.7%, while m AP only by 1.8%. The results show that the three lightweight detection networks proposed in this paper have good performance in UAV aerial image object detection task.

关键词： Training YOLO Deconvolution Accuracy Object detection Autonomous aerial vehicles Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Selecting matching area based on deep learning keypoint features 12

Selecting matching area based on deep learning keypoint feat...

引用

SPIE 12th International Symposium on multispectral Image processing and Pattern Recognition, MIPPR 2023

作者： Ding, Tian Tian, Tian Tian, Jinwen National Key Laboratory of Multispectral Information Intelligent Processing Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9781510674943

ISBN: (纸本)9781510674936

Selecting the appropriate matching area is one of the crucial tasks in scene matching systems, serving as the basis for trajectory planning. Currently, there are two main approaches for adaptation zone selection: Traditional statistical feature-based adaptation zone selection methods and deep learning-based adaptation zone selection methods. The former relies on multi-Threshold screening to choose adaptation zones, resulting in a complex and less accurate process. The latter employs model training to select adaptation zones, improving accuracy, but it tends to yield less stable results. Therefore, this paper utilizes the *** network to extract keypoints and their feature descriptors within the scene region as the initial feature set. It employs support vector regression as the matching probability prediction model and uses threshold segmentation on the predicted probabilities to ultimately determine the adaptability of the region. Experimental results demonstrate that the adaptation zone selection method proposed in this paper offers higher accuracy and stability. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Navigation

来源：评论

学校读者我要写书评

暂无评论

An Adaptive Parallel Gaussian Homotopy Approach for Zeroth-Order Optimization

An Adaptive Parallel Gaussian Homotopy Approach for Zeroth-O...

引用

2024 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2024

作者： Zhou, Bingyi Wang, Guoyou Huazhong University of Science and Technology National Key Laboratory of Multispectral Information Intelligent Processing Technology Wuhan China

ISBN: (纸本)9781665410205

The Gaussian homotopy method is a classical optimization approach for solving nonconvex problems. It applies Gaussian smoothing to a given problem, using varying smoothing factors to generate a series of proxy problems that are easier to solve, subsequently addressing them gradually from simpler to more complex ones. Traditional Gaussian homotopy methods typically utilize a predefined sequence or parameter to update a single smoothing factor serially. However, this sequential updating process is inefficient and unsuitable for parallel homotopy pipelines. Moreover, the utilization of predefined sequences or parameters leads to a lack of adaptability. To address the abovementioned challenges, we propose an adaptive parallel Gaussian homotopy optimization method. Initially, we introduce an adaptive model called Scaling. This model is formulated as a scaler that concurrently adjusts multiple smoothing factors, with each factor aligning to a homotopy level. It can be integrated easily into serial and parallel homotopy pipelines. Furthermore, we establish a collaborative training regimen to jointly train the Scaling model and a parallel homotopy model, named continuation path learning (CPL) [1] model. Throughout the training process, the Scaling model furnishes CPL with multiple scaled smoothing factors and updates CPL implicitly. Extensive experiments demonstrate that the proposed Gaussian homotopy approach performs competitively. © 2024 IEEE.

关键词： Gaussian distribution

来源：评论

学校读者我要写书评

暂无评论

An anti-occlusion tracking algorithm based on mask strategy and adaptive response fusion 12

An anti-occlusion tracking algorithm based on mask strategy ...

引用

SPIE 12th International Symposium on multispectral Image processing and Pattern Recognition, MIPPR 2023

作者： Duan, Zhongyu Huang, Shixiang Yue, Zhongmu Tian, Tian Tian, Jinwen National Key Laboratory of Multispectral Information Intelligent Processing Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9781510674943

ISBN: (纸本)9781510674936

In long-Term target tracking, target occlusion is an important factor that affects the performance of the tracker, and it is also an inevitable problem in long-Term target tracking. Therefore, it is necessary to adopt corresponding strategies for target occlusion to improve the robustness of the tracker in corresponding scenarios. In response to the problem of target loss in complex background scenarios such as target occlusion in current filtering based tracking algorithms, this paper studies the use of masking strategies and adaptive response fusion methods to improve the accuracy and robustness of tracking *** experimental results demonstrate the effectiveness of various strategies in resisting occlusion in tracking algorithms. The experimental results of the overall algorithm indicate that the algorithm can effectively solve occlusion problems while balancing accuracy, speed, and robustness. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Target tracking

来源：评论

学校读者我要写书评

暂无评论

The blessing of Depth Anything:An almost unsupervised approach to crop segmentation with depth-informed pseudo labeling

引用

植物表型组学（英文） 2025年第1期7卷 49-65页

作者： Songliang Cao Binghui Xu Wei Zhou Letian Zhou Jiafei Zhang Yuhui Zheng Weijuan Hu Zhiguo Han Hao Lu National Key Laboratory of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhan430074China PhenoTrait Technology Co. Ltd.Beijing100096China MetaPheno Laboratory Shanghai201114China National Key Laboratory of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhan430074China PhenoTrait Technology Co. Ltd.Beijing100096China MetaPheno Laboratory Shanghai201114China SpeCloud Technology Co. Ltd.Sanya572025China State Key Laboratory of Plant Cell and Chromosome Engineering Institute of Genetics and Developmental BiologyChinese Academy of SciencesBeijing100101China

We present Depth-Informed Crop Segmentation(DepthCropSeg),an almost unsupervised crop segmentation approach without manual pixel-level *** segmentation is a fundamental vision task in agriculture,which benefits a number of downstream applications such as crop growth monitoring and yield *** the past decade,image-based crop segmentation approaches have shifted from classic color-based paradigms to recent deep learning-based *** latter,however,rely heavily on large amounts of data with high-quality manual annotation such that considerable human labor and time are *** this work,we leverage Depth Anything V2,a vision foundation model,to produce high-quality pseudo crop masks for training segmentation *** compile a dataset of 17,199 images from six public plant segmentation sources,generating pseudo masks from depth maps after normalization and *** a coarse-to-fine manual screening,1378 images with reliable masks are *** compare four semantic segmentation models and enhance the top-performing one with depth-informed two-stage self-training and depth-informed *** evaluate the feasibility and robustness of DepthCropSeg,we benchmark the segmentation performance on 10 public crop segmentation testing sets and a self-collect dataset covering in-field,laboratory,and unmanned aerial vehicle(UAV)*** results show that our DepthCropSeg approach can achieve crop segmentation performance comparable to the fully supervised model trained with manually annotated data(86.91 vs.87.10).For the first time,we demonstrate almost unsupervised,close-to-full-supervision crop segmentation successfully.

关键词： Crop segmentation Plant phenotyping Depth anything Segment anything Efficient labeling

来源：评论

学校读者我要写书评

暂无评论

High-Throughput and Energy-Efficient FPGA-Based Accelerator for All Adder Neural Networks

引用

IEEE Internet of Things Journal 2025年第12期12卷 20357-20376页

作者： Zhang, Ning Ni, Shuo Chen, Liang Wang, Tong Chen, He Beijing Institute of Technology National Key Laboratory of Science and Technology on Space-Born Intelligent Information Processing Beijing100081 China

Neural networks have been extensively applied across various Internet of Things (IoT) applications, such as drone-and satellite-based remote sensing and autonomous driving. With the increasing resolution and amount of data captured by sensors, the demand for real-time response in IoT applications is markedly increasing. However, it is difficult for existing convolutional neural network (CNN) accelerators for IoT applications on field-programmable gate array (FPGA) platforms to achieve high throughput because of the inherent dense multiplication operations of CNNs, memory bandwidth limitations and inefficient mapping mechanisms. In this article, a high-throughput and energy-efficient all adder neural network (ANN) accelerator for IoT applications on FPGA platform is proposed to solve this problem. First, a series of hardware-oriented algorithm optimization methods are proposed to simplify the processing flow of ANN and further minimize its deployment overhead. Second, a novel hardware architecture based on the idea of near-memory computation (NMC) is proposed to eliminate off-chip memory access completely and accelerate the reconstructed ANN in the pipeline. Third, a set of quantitative analysis methods for the proposed accelerator is presented to balance throughput and energy consumption, allowing the accelerator to adapt to the varying demands of different IoT application scenarios. Extensive experimental results on the AMD-Xilinx VC709 board demonstrate that the proposed accelerator achieves state-of-the-art performance in terms of throughput, energy efficiency, and throughput efficiency. Moreover, experiments on the AMD-Xilinx KV260 board highlight the architecture's exceptional scalability and energy efficiency, enabling a balance between speed and power consumption tailored to the specific requirements of IoT application scenarios. © 2014 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Semantic segmentation via pixel-to-center similarity calculation

引用

CAAI Transactions on Intelligence technology 2024年第1期9卷 87-100页

作者： Dongyue Wu Zilin Guo Aoyan Li Changqian Yu Nong Sang Changxin Gao National Key Laboratory of Science and Technology on Multispectral Information Processing School of Artificial Intelligence and AutomationHuazhong University of Science and TechnologyWuhanChina Meituan Inc. BeijingChina

Since the fully convolutional network has achieved great success in semantic segmentation,lots of works have been proposed to extract discriminative pixel ***,the authors observe that existing methods still suffer from two typical challenges:(i)The intra-class feature variation between different scenes may be large,leading to the difficulty in maintaining the consistency between same-class pixels from different scenes;(ii)The inter-class feature distinction in the same scene could be small,resulting in the limited performance to distinguish different classes in each *** authors first rethink se-mantic segmentation from a perspective of similarity between pixels and class *** weight vector of the segmentation head represents its corresponding semantic class in the whole dataset,which can be regarded as the embedding of the class ***,the pixel-wise classification amounts to computing similarity in the final feature space between pixels and the class *** this novel view,the authors propose a Class Center Similarity(CCS)layer to address the above-mentioned challenges by generating adaptive class centers conditioned on each scenes and supervising the similarities between class *** CCS layer utilises the Adaptive Class Center Module to generate class centers conditioned on each scene,which adapt the large intra-class variation between different *** designed Class Distance Loss(CD Loss)is introduced to control both inter-class and intra-class distances based on the predicted center-to-center and pixel-to-center ***,the CCS layer outputs the processed pixel-to-center similarity as the segmentation *** experiments demonstrate that our model performs favourably against the state-of-the-art methods.

关键词： computer vision deep neural networks image segmentation scene understanding

来源：评论

学校读者我要写书评

暂无评论

Self-distilled Dual-Network with Pixel Screening Loss for Blind Image Deblurring 27th

Self-distilled Dual-Network with Pixel Screening Loss for ...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Li, Tianyi Tian, Ming Gao, Changxin Sang, Nong National Key Laboratory of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9783031781247

In the field of image processing, blind image deblurring aims to restore sharp details in images blurred by an unknown convolution kernel. Recent advancements have shown that deep networks can act as effective image generative priors (DIP) for restoring clear images without requiring external datasets. However, the inherent non-uniqueness of solutions in blind image deblurring often leads DIP-based methods to converge on local optima, resulting in over- or under-deblurred images. To overcome this limitation, we propose a novel deblurring framework featuring dual image generators. These generators mutually constrain each other during training, guiding the model towards the optimal solution. Building on our network structure, we employ a self-ensemble and self-distillation strategy to guide network training, further enhancing performance. Additionally, we introduce a novel loss function based on a pixel screening method, which focuses on the important pixels. This loss enables the network to model the blur kernel more accurately and facilitates the restoration of image details. Our experiments demonstrate that our deblurring approach outperforms most existing methods both qualitatively and quantitatively. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Restoration

来源：评论

学校读者我要写书评

暂无评论

Robust facial landmark detection network based on multiscale attention residual blocks 12

Robust facial landmark detection network based on multiscale...

引用

SPIE 12th International Symposium on multispectral Image processing and Pattern Recognition, MIPPR 2023

作者： Zou, Lamei Xie, Jia Wang, Hao Lu, Luhan Wu, Chengqing Guo, Yichun Zhang, Tianbao National Key Laboratory of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9781510674967

ISBN: (纸本)9781510674950

We present a robust facial landmark detection network based on multiscale attention residual blocks (MARBNet) for effectively predicting facial landmark. MARBNet consists of three modules. Firstly, the coarse feature extraction module obtains coarse features through convolution, batch normalization, ReLU activation, and maximum pooling. The fine feature extraction module is composed of 33 multiscale attention residual blocks (MARB). MARB is composed of 1x1 convolution layer, 3x3 convolution layer, 1x1 convolution layer, two multiscale convolution module(MulRes) and channel attention module(CAM). MulRes is used to extract complementary features of different scales, obtain more feature information under different Receptive field, and avoid excessive loss of key information in the input image. CAM enables the network to pay more attention to high-frequency information on the channel, effectively prevents the loss of information, so as to improve the effect of facial landmark detection. The output module consists of two 1x1 convolution layers, one of which outputs landmark heatmap score and landmark coordinate offset, and the other outputs the nearest neighbor landmark offset. The experiment results on WFLW and 300W datasets show that our method is superior to the existing algorithms in terms of normalized mean square error indicators. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Complementary Dual-Branch Network for Space-Time Video Super-Resolution 27th

Complementary Dual-Branch Network for Space-Time Video Supe...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Tian, Ming Li, Tianyi Luo, RongSheng Gao, Changxin Sang, Nong National Key Laboratory of Multispectral Information Intelligent Processing Technology School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9783031781247

Space-time video super-resolution aims to simultaneously increase the space-time resolution of low-resolution and low frame-rate videos. Existing deep learning-based methods have made notable strides, predominantly achieving space-time video super-resolution through the relatively simple integration of modules for video super-resolution and video frame interpolation sub-tasks. However, these methods typically do not fully exploit the inherent relationships between the two sub-tasks. To address this limitation, we propose a Complementary Dual-Branch Network designed to better explore the interdependence of the two sub-tasks. Specifically, our dual-branch architecture facilitates mutual enhancement between video super-resolution and video frame interpolation sub-tasks within each branch and provides mutual guidance between the two branches. Additionally, we introduce a simple yet effective strategy for the rough estimation of optical flow, incorporating Flow-Guided Deformable Alignment into space-time video super-resolution to achieve precise motion estimation. In addition, we use an RNN-based Backward and Forward Recurrent module to ensure that all frames can utilize the information of the whole sequence. It is more efficient and memory saving compared to the currently popular bidirectional LSTM module. Experimental results on several datasets show that our method achieves superior accuracy and requires fewer parameters compared to state-of-the-art methods. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Optical flows

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：