检索结果-内蒙古大学图书馆

Conference on Computer Vision and pattern recognition (CVPR)

作者： Nando Metzger Rodrigo Caye Daudt Konrad Schindler Photogrammetry and Remote Sensing ETH Zurich

Performing super-resolution of a depth image using the guidance from an RGB image is a problem that concerns several fields, such as robotics, medical imaging, and remote sensing. While deep learning methods have achieved good results in this problem, recent work highlighted the value of combining modern methods with more formal frameworks. In this work, we propose a novel approach which combines guided anisotropic diffusion with a deep convolutional network and advances the state of the art for guided depth super-resolution. The edge transferring/enhancing properties of the diffusion are boosted by the contextual reasoning capabilities of modern networks, and a strict adjustment step guarantees perfect adherence to the source image. We achieve unprecedented results in three commonly used benchmarks for guided depth superresolution. The performance gain compared to other methods is the largest at larger scales, such as × 32 scaling. Code 1 1 https://***/prs-eth/Diffusion-Super-Resolution for the proposed method is available to promote reproducibility of our results.

关键词：

来源：评论

学校读者我要写书评

暂无评论

1st International Conference on pattern Analysis and Machine Intelligence, ICPAMI 2024

1st International Conference on Pattern Analysis and Machine...

引用

1st International Conference on pattern Analysis and Machine Intelligence, ICPAMI 2024

ISBN: (纸本)9789819633487

The proceedings contain 28 papers. The special focus in this conference is on pattern Analysis and Machine Intelligence. The topics include: Development of a Low Cost 3D LiDAR Using 2D LiDAR and Servo Motor;the Design of Machine Vision-Based Waste Sorting System;ECLNet: Efficient Convolution with Lite Transformer for 3D Medical image Segmentation;exploring High-Performance 3D Object Detection with Partial Depth Completion;full-Scale Network for remote sensing Object Detection;Detection of Pedestrian Movement Poses in High-Speed Autonomous Driving Environments Using DVS;city-Scale Multi-Camera Vehicle Tracking System with Improved Self-Supervised Camera Link Model;an Efficient Transformer-Based Network for remote sensing image Change Detection;the Method for Three-Dimensional Visual Measurement of Circular Markers Based on Active Fusion Technology;intelligent image recognition and Classification Technology in Digital Media;Indoor Visible Light Positioning System Based on the image Sensor and CNN-GRU Fusion Neural Network;Stock Investor Sentiment Analysis Based on NLP;Novel Audiobook System Based on BERT;student Enrollment Consultation Q&A Robot Based on Large Language Model;family Doctor Model Training Based on Large Language Model Tuning;composite Awareness-Based Knowledge Distillation for Medical Anomaly Detection;Improved CNN-GRU RF Fingerprint Feature recognition Method Based on Comb Filter;emotional State recognition of English Learners Based on Deep Learning;Application of Classification Framework Based on CDR and CNN in Ophthalmic Prediagnosis;visual recognition and Recommendation System for Cultural Tourism Attractions Based on Deep Learning;quadruped Robot System Based on Proprioceptive Vision and Complex Ground Mobility Capabilities;a Simulated Dataset to Evaluate the Visual-Inertial Odometry Algorithms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Mitigating Challenges of the Space Environment for Onboard Artificial Intelligence: Design Overview of the Imaging Payload on SpIRIT

Mitigating Challenges of the Space Environment for Onboard A...

引用

IEEE Computer Society Conference on Computer Vision and pattern recognition Workshops (CVPRW)

作者： Miguel Ortiz Del Castillo Jonathan Morgan Jack McRobbie Clint Therakam Zaher Joukhadar Robert Mearns Simon Barraclough Richard Sinnott Andrew Woods Chris Bayliss Kris Ehinger Ben Rubinstein James Bailey Airlie Chapman Michele Trenti The University of Melbourne

ISBN: (数字)9798350365474

ISBN: (纸本)9798350365481

Artificial intelligence (AI) and autonomous edge computing in space are emerging areas of interest to augment capabilities of nanosatellites, where modern sensors generate orders of magnitude more data than can typically be transmitted to mission control. Here, we present the hardware and software design of an onboard AI subsystem hosted on SpIRIT. The system is optimised for on-board computer vision experiments based on visible light and long wave infrared cameras. This paper highlights the key design choices made to maximise the robustness of the system in harsh space conditions, and their motivation relative to key mission requirements, such as limited compute resources, resilience to cosmic radiation, extreme temperature variations, distribution shifts, and very low transmission bandwidths. The payload, called Loris, consists of six visible light cameras, three infrared cameras, a camera control board and a Graphics processing Unit (GPU) system-on-module. Loris enables the execution of AI models with on-orbit fine-tuning as well as a next-generation image compression algorithm, including progressive coding. This innovative approach not only enhances the data processing capabilities of nanosatellites but also lays the groundwork for broader applications to remote sensing from space.

关键词： Computer vision Temperature distribution image coding Space missions Small satellites Graphics processing units Aerospace electronics

来源：评论

学校读者我要写书评

暂无评论

How Do Human Detect Targets of remote sensing images with Visual Attention?

SSRN

引用

SSRN 2024年

作者： He, Bing Qin, Tong Shi, Bowen Dong, Weihua Ghent University Belgium

Human visual attention is the basis of target recognition, change detection and classification in remote sensing images. However, the human visual attention of remote sensing images during target detection remains uninvestigated. In this study, we simultaneously collected eye tracking and electroencephalography (EEG) data of 40 experts during target detection in 1000 remote sensing images. We quantified their attention in different stages of target detection (i.e., target search, selection, and identification) from their eye movements and brain activities. The eye-tracking results indicated that hue and brightness were crucial visual features that guided visual search for remote sensing images. The size of the target also affected the allocation of human attention resources. Particularly, participants tended to miss the targets that were smaller than 4% of the whole image area. The fixation event-related potentials (FRPs) in temporal and parietal brain regions further distinguished the attention of targets and distractors, which could be attributed to the process of memory and processing. Our findings offer new visual and neural evidence for human attention in target detection of remote sensing images, which not only contribute to the performance of attention-based algorithms in remote sensing interpretation, but also provide an empirical basis for guiding human-machine intelligent fusion. © 2024, The Authors. All rights reserved.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

基于TransMANet的遥感图像语义分割算法

引用

激光与光电子学进展 2024年第10期61卷 301-312页

作者：宋熙睿葛洪伟江南大学人工智能与计算机学院江苏无锡214122 江苏省模式识别与计算智能工程实验室(江南大学) 江苏无锡214122

针对multiattention network(MANet)算法与图像语义信息关联不足、全局特征提取不充分和分割精度较低的问题,基于Transformer与注意力机制,提出一种增强浅层网络语义信息,具有融合局部和全局上下文的双分支解码器的网络结构,即Transform... 详细信息

针对multiattention network(MANet)算法与图像语义信息关联不足、全局特征提取不充分和分割精度较低的问题,基于Transformer与注意力机制,提出一种增强浅层网络语义信息,具有融合局部和全局上下文的双分支解码器的网络结构,即Transformer multiattention network(TransMANet)。首先,引入局部注意力嵌入机制,增强上下文信息的嵌入,并将高级特征的语义信息嵌入低级特征;然后,设计基于Transformer与卷积神经网络的双分支解码器,分别提取全局上下文信息和不同尺度的细节信息,对全局与局部信息建模;最后,改进原有的损失函数,缓解遥感数据集类别不平衡的问题,提高分割准确度。实验结果表明,TransMANet在UAVid、LoveDA、Potsdam和Vaihingen数据集上均取得了较MANet及其他有竞争力的先进方法更优的交并比指标,有较好的泛化能力。

关键词：图像处理语义分割注意力机制 Transformer 高分辨率遥感影像

来源：评论

学校读者我要写书评

暂无评论

Content-Adaptive Non-Local Convolution for remote sensing Pansharpening

Content-Adaptive Non-Local Convolution for Remote Sensing Pa...

引用

Conference on Computer Vision and pattern recognition (CVPR)

作者： Yule Duan Xiao Wu Haoyu Deng Liang-Jian Deng University of Electronic Science and Technology of China China

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Currently, machine learning-based methods for remote sensing pansharpening have progressed rapidly. However, existing pansharpening methods often do not fully exploit differentiating regional information in non-local spaces, thereby limiting the effectiveness of the methods and resulting in redundant learning parameters. In this pa-per, we introduce a socalled content-adaptive non-local convolution (CANConv), a novel method tailored for re-mote sensing image pansharpening. Specifically, CANConv employs adaptive convolution, ensuring spatial adaptability, and incorporates non-local self-similarity through the similarity relationship partition (SRP) and the partition-wise adaptive convolution (PWAC) sub-modules. Furthermore, we also propose a corresponding network architecture, called CANNet, which mainly utilizes the multi-scale self-similarity. Extensive experiments demonstrate the superior performance of CANConv, compared with recent promising fusion methods. Besides, we substantiate the method's effectiveness through visualization, ablation experiments, and comparison with existing methods on multiple test sets. The source code is publicly available at https://***/duany11/CANConv.

关键词： Learning systems Visualization Limiting Convolution Source coding Pansharpening Network architecture

来源：评论

学校读者我要写书评

暂无评论

Graphical User Interface Development with Database Integration for Cal-Val Analysis of Multiple SAR Missions 9

Graphical User Interface Development with Database Integrati...

引用

9th IEEE International Conference for Convergence in Technology, I2CT 2024

作者： Basak, Aakashneel Niharika, K. Jayasri, P.V. Usha Sundari Ryali, H.S.V. Sarma, Manju Indian Space Research Organisation Sar Data Processing Division Microwave Data Processing Group Data Processing Area National Remote Sensing Centre India

ISBN: (纸本)9798350394474

Synthetic Aperture Radar (SAR) Calibration & Validation are critical processes to ensure the accuracy, precision and reliability of SAR imaging systems. These activities are particularly essential in dynamic environments over extended periods, where changes in onboard SAR payload, data processing software or environmental conditions can influence the system performance. This paper provides a detailed exposition on the conceptualization and development of a Graphical User Interface (GUI), specifically designed to input essential parameters for comprehensive assessment of radiometric, geometric and polarimetric SAR image quality parameters from both Point and Distributed target data. The software excels in presenting analysis results in a well-organized tabulated format & offers a valuable feature allowing users to have the visual representation of the Impulse Response and Gamma Naught pattern data. A standout attribute of the software lies in its ability to effortlessly store and retrieve analysis data through its interface with SQLite Database. This seamlessly integrated Database System streamlines the organization and archival of calibration results, establishing a thorough historical record for each analysis. The utilization of the SQLite database guarantees data integrity & empowers users to effortlessly retrieve and compare calibration data over time, fostering effective trend analysis and enabling a comprehensive evaluation of the long-term performance of SAR system. The software has been created with Python programming language as front-end and SQLite as part of back-end. The developed GUI software package represents a significant advancement in SAR system calibration methodologies, providing SAR data analyst with a user-friendly interface for efficient SAR calibration analysis. © 2024 IEEE.

关键词： Synthetic aperture radar

来源：评论

学校读者我要写书评

暂无评论

GeoChat:Grounded Large Vision-Language Model for remote sensing

GeoChat:Grounded Large Vision-Language Model for Remote Sens...

引用

Conference on Computer Vision and pattern recognition (CVPR)

作者： Kartik Kuckreja Muhammad Sohail Danish Muzammal Naseer Abhijit Das Salman Khan Fahad Shahbaz Khan Mohamed bin Zayed University of AI Birla Institute of Technology & Science Hyderabad Australian National University Linkoping University

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Recent advancements in Large Vision-Language Models (VLMs) have shown great promise in natural image domains, allowing users to hold a dialogue about given visual content. However, such general-domain VLMs perform poorly for remote sensing (RS) scenarios, leading to inaccurate or fabricated information when presented with RS domain-specific queries. Such a behavior emerges due to the unique challenges introduced by RS imagery. For example, to handle high-resolution RS imagery with diverse scale changes across categories and many small objects, region-level reasoning is necessary alongside holistic scene inter-pretation. Furthermore, the lack of domain-specific multimodal instruction following data as well as strong back-bone models for RS make it hard for the models to align their behavior with user queries. To address these limitations, we propose GeoChat - the first versatile remote sensing VLM that offers multitask conversational capabilities with high-resolution RS images. Specifically, GeoChat can not only answer image-level queries but also accepts region inputs to hold region-specific dialogue. Further-more, it can visually ground objects in its responses by referring to their spatial coordinates. To address the lack of domain-specific datasets, we generate a novel RS multimodal instruction-following dataset by extending image-text pairs from existing diverse RS datasets. We establish a comprehensive benchmarkfor RS multitask conversations and compare with a number of baseline methods. GeoChat demonstrates robust zero-shot performance on various RS tasks, e.g., image and region captioning, visual question answering, scene classification, visually grounded conversations and referring detection. Our code is available here.

关键词： Visualization Scene classification Grounding Oral communication Object detection Benchmark testing Data models

来源：评论

学校读者我要写书评

暂无评论

Long-Tailed SAR Target recognition Based on Expert Network and Intraclass Resampling

引用

IEEE GEOSCIENCE AND remote sensing LETTERS 2023年 20卷 1-1页

作者： Liu, Yingbing Zhang, Fan Ma, Lixiang Ma, Fei Beijing Univ Chem Technol Sch Informat Sci & Technol Beijing 100029 Peoples R China China Acad Space Technol CAST Inst Remote Sensing Satellite Beijing 100094 Peoples R China

In recent years, it has been a research hotspot to apply big data-driven deep learning methods to synthetic aperture radar (SAR) target recognition with limited data. However, the problem caused by the long-tailed characteristics of SAR data has long been ignored. Specifically, a majority of data samples are concentrated in a few categories, leading to a skewed distribution of data. This skewed distribution can cause learning bias toward the majority class, which can subsequently degrade the recognition performance of the minority class. This issue is further exacerbated in limited sample conditions for SAR target recognition. After conducting research on target recognition for long-tailed natural images, this study has found that the existing methods used in this field cannot be easily applied to SAR target recognition. The primary reason is that SAR image data exhibit simultaneous and complex interclass and intraclass long-tailed distributions. In response to this issue, we propose the use of a multibranch expert network and dual-environment sampling to address the long-tail problems in both interclass and intraclass scenarios. The proposed method outperforms popular long-tailed target recognition methods on the long-tailed versions of the MSTAR and FUSAR datasets.

关键词： Long-tailed distributions synthetic aperture radar (SAR) target recognition

来源：评论

学校读者我要写书评

暂无评论

QFabric: Multi-Task Change Detection Dataset

QFabric: Multi-Task Change Detection Dataset

引用

IEEE/CVF Conference on Computer Vision and pattern recognition (CVPR)

作者： Verma, Sagar Panigrahi, Akash Gupta, Siddharth Univ Paris Saclay INRIA Cent Supelec Ctr Vis Numer Paris France Granular AI Somerville NJ USA

ISBN: (纸本)9781665448994

Detecting change through multi-image, multi-date remote sensing is essential to developing and understanding of global conditions. Despite recent advancements in remote sensing realized through deep learning, novel methods for accurate multi-image change detection remain unrealized. Recently, several promising methods have been proposed to address this topic, but a paucity of publicly available data limits the methods that can be assessed. In particular, there exists limited work on categorizing the nature and status of change across an observation period. This paper introduces the first labeled dataset available for such a task. We present an open-source change detection dataset, termed QFabric, with 450,000 change polygons annotated across 504 locations in 100 different cities covering a wide range of geographies and urban fabrics. QFabric is a temporal multi-task dataset with 6 change types and 9 change status classes. The geography and environment metadata around each polygon provides context that can be leveraged to build robust deep neural networks. We apply multiple benchmarks on our dataset for change detection, change type and status classification tasks. Project page: https://***/qfabric

关键词： Deep learning Urban areas Predictive models Benchmark testing Metadata Fabrics pattern recognition

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：