检索结果-内蒙古大学图书馆

4th International conference on Multimedia processing, communication and information Technology, MPCIT 2024

作者： Vandana, N. vidyasagar, K.N. Suguresh Hanamanal, Vachan Arun Pujari, Rahul R. School of Electronics and Communication Engineering REVA University India

ISBN: (纸本)9798350375466

The proposed system integrates YOLOv5 [1] for object detection, CMFNet [2] for image enhancement, and ResNet [3] for image classification into a unified framework for comprehensive image analysis. This integrated approach overcomes limitations of standalone methods, providing a seamless platform for various image analysis tasks. The system demonstrates improved performance in visual quality enhancement, real-time object detection, and accurate image classification across diverse scenarios. The proposed design allows for flexibility and future extensibility, making it adaptable to evolving requirements in computer vision applications. © 2024 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Improved non-local GBT model for depth map denoising 4

Improved non-local GBT model for depth map denoising

引用

4th International conference on Signal Image processing and communication, ICSIPC 2024

作者： Meng, Yaqun Ge, Huayong Hou, Xinxin Ji, Yukai College of Information Science and Technology Donghua University Shanghai261020 China

ISBN: (纸本)9781510682467

The non-local GBT model (NLGBT) shows excellent performance in terms of speed in depth map denoising. However, this model suffers from more artifacts and difficulty in recovering image details. Therefore, in this paper, we propose an improved NLGBT model. Specifically, we utilize an improved noise iteration update parameter supplemented with a non-fixed number of iterations, sacrificing a small amount of denoising speed to greatly improve the denoising accuracy. Experiments show that the improved NLGBT model not only denoises quickly, but also suppresses artifacts while preserving the local structure of the image. © 2024 SPIE.

关键词： Denoising Depth maps Signal to noise ratio Performance modeling Image processing Image denoising Image enhancement Tunable filters visual process modeling visualization

来源：评论

学校读者我要写书评

暂无评论

Poster Abstract: Joint Optical Wireless communication and Sensing using Neuromorphic Cameras 23

Poster Abstract: Joint Optical Wireless Communication and Se...

引用

23rd ACM/IEEE International conference on information processing in Sensor Networks, IPSN 2024

作者： Mohamed Nishar, Abbaas Alif Paul, Sonipriya Ashok, Ashwin Georgia State Univerisity AtlantaGA United States

ISBN: (纸本)9798350362015

In this work, we propose a novel re-use of neuromorphic (event) cameras for joint sensing and communications. Event cameras work on the principle of capturing changes in the light intensities, essentially capturing events that lead to such changes. This makes them operable at low power and sample events at fast rates (equivalent to about 40K frames-per-second compared to RGB cameras). We propose a system design to leverage the time-sampling nature of events for optical wireless communication and the ability to sample a collective area of physical space for imaging. In particular, we propose to address the challenges to achieve passive optical wireless (backscatter) communication as well as computer vision functions such as object and path detection using a single neuromorphic camera device. We posit that such an integrated functioning through a single low-power device opens new avenues for visible/invisible light communication and visual scene *** CONCEPTS•Networks → Mobile networks;•Hardware → Signal processing systems;•Computing methodologies → Computer vision. © 2024 IEEE.

关键词： Cameras

来源：评论

学校读者我要写书评

暂无评论

Translation Tool for Alternative Communicators using Natural Language processing 5

Translation Tool for Alternative Communicators using Natural...

引用

5th International conference on Electronics and Sustainable communication Systems, ICESC 2024

作者： Fareed, A. Irfaan Ramanathan, M. Yeswanth, R. Devi, S. Coimbatore Institute of Technology B-Tech Information Technology Tamilnadu Coimbatore India

ISBN: (纸本)9798350379945

Accessibility and integration are essential in today's world, but people with visual, hearing, and speech impairments often face significant communication barriers. Historically, creating accessible formats like Braille for the blind or transcribing sign language for the deaf has been time-consuming and bureaucratic. Individuals with disabilities have had to deal with long wait times and a heavy reliance on external resources to access information in their preferred formats. The proposed system offers a comprehensive solution by introducing a device that instantly converts various forms of input into accessible formats, bridging communication gaps between individuals with visual, auditory, and speech impairments. This device can convert typed text into Braille, audio, or text for those who are blind, deaf, or mute. It can also translate sign language gestures into text, Braille, or audio, enhancing communication. The most important innovation of the proposed system is its ability to handle multiple input and output formats. Users no longer need to navigate the bureaucracy of transcribing documents. They can enter text, speak into the device, or use sign language gestures to receive instant translations. This includes converting text to audio, sign language to text, sign language to Braille, and audio to Braille. This revolutionary process leads to greater freedom and access to information for individuals with visual, auditory, and speech impairments, promoting inclusivity and ensuring equal access to technology for all. © 2024 IEEE.

关键词： Audition

来源：评论

学校读者我要写书评

暂无评论

CLIP in Mirror: Disentangling text from visual images through reflection 38

CLIP in Mirror: Disentangling text from visual images throug...

引用

38th conference on Neural information processing Systems, NeurIPS 2024

作者： Wang, Tiancheng Yang, Yuguang Yang, Linlin Lin, Shaohui Zhang, Juan Guo, Guodong Zhang, Baochang Institute of Artificial Intelligence Beihang University Beijing China School of Electronic Information Engineering Beihang University Beijing China Zhongguancun Laboratory Beijing China State Key Laboratory of Media Convergence and Communication Communication University of China Beijing China School of Computer Science and Technology East China Normal University Shanghai China Ningbo Institute of Digital Twin Eastern Institute of Technology Ningbo China

The CLIP network excels in various tasks, but struggles with text-visual images i.e., images that contain both text and visual objects;it risks confusing textual and visual representations. To address this issue, we propose MirrorCLIP, a zero-shot framework, which disentangles the image features of CLIP by exploiting the difference in the mirror effect between visual objects and text in the images. Specifically, MirrorCLIP takes both original and flipped images as inputs, comparing their features dimension-wise in the latent space to generate disentangling masks. With disentangling masks, we further design filters to separate textual and visual factors more precisely, and then get disentangled representations. Qualitative experiments using stable diffusion models and class activation mapping (CAM) validate the effectiveness of our disentanglement. Moreover, our proposed MirrorCLIP reduces confusion when encountering text-visual images and achieves a substantial improvement on typographic defense, further demonstrating its superior ability of disentanglement. Our code is available at https://***/tcwangbuaa/MirrorCLIP. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fake Logo Detection Using Image processing

Fake Logo Detection Using Image Processing

引用

2024 IEEE International conference on information Technology, Electronics and Intelligent communication Systems, ICITEICS 2024

作者： Battini, Vengamamba Sravya Kishan, S. Ravi Valeti, Venkata Divya Vrsec Dept. of Cse Andhra Pradesh Kanuru India

ISBN: (纸本)9798350382693

Logo authentication is crucial in various domains to ensure the integrity and trustworthiness of visual brand representations. This project introduces a sophisticated Fake Logo Detection system leveraging Convolutional Neural Networks (CNNs) to ascertain the legitimacy of provided logos by conducting a thorough analysis of key features, visual attributes, and patterns. In the contemporary landscape, logos play a pivotal role in defining a company's identity, and the proliferation of counterfeit logos poses a substantial threat to brand credibility. The innovative CNN-based methodology exploits its inherent capability to autonomously extract intricate features and patterns from logo images, facilitating a robust differentiation between authentic and manipulated logos. The model undergoes comprehensive training on a diverse dataset that spans a wide spectrum of logos, ensuring its adaptability to various visual styles. Through a meticulous examination of features, visual characteristics, and patterns, our system contributes to a nuanced comprehension of logo authenticity. Empirical results underscore the effectiveness of the proposed approach in accurately discerning real and fake logos, thus providing a valuable tool for safeguarding brands in the digital domain. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Research on Ai Technology of Intelligent information visualization Based on Rtos System Service 2

Research on Ai Technology of Intelligent Information Visuali...

引用

2nd International conference on Smart Energy, ICSE 2023

作者： Jinglu, Zhang Linna, Wang Beijing Polytechnic Beijing100176 China

In order to improve the visual interaction ability of intelligent information, an intelligent information visualization artificial intelligence technology based on RTOS system service is proposed in this paper. Based on embedded integrated control, this paper constructs the system design scheme, client and host system of intelligent information visualization artificial intelligence RTOS service system. The information monitoring module, artificial intelligence programmable logic control module, artificial intelligence processing module and human-computer interaction module of intelligent information visualization artificial intelligence RTOS service system are established to realize the label identification and identification in the process of intelligent information visualization artificial intelligence RTOS service system based on unified resource locator sipur, The communication system model of intelligent information visualization artificial intelligence RTOS service system under the transmission mode of multimedia communication channel is established to realize the automatic information processing of intelligent information visualization RTOS system. The test results show that the design has good stability, strong intelligent information visualization and interaction ability, low memory and time cost. © Published under licence by IOP Publishing Ltd.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Research on Graphic processing System of visual communication Design Based on Computer Big Data

Research on Graphic Processing System of Visual Communicatio...

引用

2023 IEEE International conference on Control, Electronics and Computer Technology, ICCECT 2023

作者： Xia, Zhenjie Li, Min Wuhan University of Bioengineering Wuhan China

ISBN: (纸本)9798350399028

In this paper, the computer image visual communication system is designed and the function of this system is analyzed. Then, based on the hardware design, this paper re-confirms the position, proportion and Angle information of 3D human body. Machine learning algorithm is introduced to integrate visual experience data with user data. This completes the filling of visual experience data. Finally, the texture information to be rendered is transformed into rendered texture by texture mapping method. Complete 3D animation rendering based on the difference between input points pixels. Finally, the design of computer graphics visual communication system is tested. The test shows that this system has powerful drawing function and can be used in many occasions. © 2023 IEEE.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

Extraction of visual information to predict crowdfunding success

引用

PRODUCTION AND OPERATIONS MANAGEMENT 2023年第12期32卷 4172-4189页

作者： Blanchard, Simon J. Noseworthy, Theodore J. Pancer, Ethan Poole, Maxwell Georgetown Univ McDonough Sch Business Washington DC USA York Univ Schulich Sch Business N York ON Canada St Marys Univ Sobey Sch Business Halifax NS Canada Georgetown Univ Washington DC 20057 USA

Researchers have increasingly turned to crowdfunding platforms to gain insights into entrepreneurial activity and dynamics. While previous studies have explored various factors influencing crowdfunding success, such as technology, communication, and marketing strategies, the role of visual elements that can be automatically extracted from images has received less attention. This is surprising, considering that crowdfunding platforms emphasize the importance of attention-grabbing and high-resolution images, and previous research has shown that image characteristics can significantly impact product evaluations. Indeed, a comprehensive review of empirical articles (n = 202) utilized Kickstarter data, focusing on the incorporation of visual information in their analyses. Our findings reveal that only 29.70% controlled for the number of images, and less than 12% considered any image details. In this manuscript, we contribute to the existing literature by emphasizing the significance of visual characteristics as essential variables in empirical investigations of crowdfunding success. We review the literature on image processing and its relevance to the business domain, highlighting two types of visual variables: visual counts (number of pictures and number of videos) and image details. Building upon previous work that discussed the role of color, composition, and figure-ground relationships, we introduce visual scene elements that have not yet been explored in crowdfunding, including the number of faces, the number of concepts depicted, and the ease of identifying those concepts. To demonstrate the predictive value of visual counts and image details, we analyze Kickstarter data using flexible machine learning models (Lasso, Ridge, Bayesian additive regression trees, and eXtreme Gradient Boosting). Our results highlight that visual count features are two of the top three predictors of success and highlight the ease at which researchers can incorporate some information abou

关键词： computer vision crowdfunding image characteristics new ventures visual scene processing

来源：评论

学校读者我要写书评

暂无评论

Extending an Industrial Robot with Image processing System 4th

Extending an Industrial Robot with Image Processing System

引用

4th International conference on Vehicle and Automotive Engineering (VAE)

作者： Lenart, Jozsef Univ Miskolc H-3515 Egyet Varos Miskolc Hungary

ISBN: (纸本)9783031152115;9783031152108

Today, industrial robots play a significant role in the automotive industry. There are several workflows in the industry that can only be solved efficiently and productively with robots. These robots are equipped with various accessories, such as grippers, sensors and, if necessary, an image processing system. There are tasks where traditional robot programming with fixed coordinates is not enough;visual information about the workspace and workpiece is needed. The problem, especially in smaller companies, is that the existing robot needs to be upgraded, for example, with a visual sensor and an image processing system, but this is usually difficult and costly for older robots and robot controllers. The aim of my research is to create a solution that allows a robot controller to be economically equipped with an image processing system that is not prepared for this at the factory. All that is required for the controller to be able to communicate with the outside world via some communication interface. An external image processing system, which can be a standard PC with the appropriate camera and software, is required. The image processing software running on the PC communicates with the robot controller, and the robot program receives the necessary information from the image processing system via the communication channel.

关键词： Robotics Machine vision Image processing

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：