检索结果-内蒙古大学图书馆

Frequency disentangled residual network

Multimedia Systems 2024年第1期30卷 9-9页

作者： Singh, Satya Rajendra Yedla, Roshan Reddy Dubey, Shiv Ram Sanodiya, Rakesh Kumar Chu, Wei-Ta Computer Science and Engineering Group Indian Institute of Information Technology Andhra Pradesh Sri India Stony Brook University Stony BrookNY United States Computer Vision and Biometrics Lab Indian Institute of Information Technology Uttar Pradesh Allahabad India Department of Computer Science and Information Engineering National Cheng Kung University Tainan Taiwan

Residual networks (ResNets) have been utilized for various computer vision and image processing applications. The residual connection improves the training of the network with better gradient flow. A residual block consists of a few convolutional layers having trainable parameters, which leads to overfitting. Moreover, the present residual networks are not able to utilize the high- and low-frequency information suitably, which also challenges the generalization capability of the network. In this paper, a frequency-disentangled residual network (FDResNet) is proposed to tackle these issues. Specifically, FDResNet includes separate connections in the residual block for low- and high-frequency components, respectively. Basically, the proposed model disentangles the low- and high-frequency components to increase the generalization ability. Moreover, the computation of low- and high-frequency components using fixed filters further avoids the overfitting. The proposed model is tested on benchmark CIFAR-10/100, Caltech, and TinyImageNet datasets for image classification. The performance of the proposed model is also tested in the image retrieval framework. It is noticed that the proposed model outperforms its counterpart residual model. The effect of kernel size and standard deviation is also evaluated. The impact of the frequency disentangling is also analyzed using a saliency map. © 2024, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Facial Emotion Recognition based on Attentive Neural Network for the Blind

Facial Emotion Recognition based on Attentive Neural Network...

引用

International Conference on Control Communication & Computing India (ICCC)

作者： Maya Dinesh Sreeni K. G. Anurenjan P. R. Computer Vision Lab College of Engineering Trivandrum Kerala APJ Abdul Kalam Technological University Trivandrum Kerala India Rajiv Gandhi Institute of Technology Kottayam Kerala

Facial expressions are considered to be a universal form of nonverbal communication. They convey countless emotions without even saying a word. These cues support speech by assisting the listener in deciphering the intended meaning of the words spoken. Unfortunately, due to the absence of access to such signs for the blind and visually impaired people (VIPs), interpersonal communication is hampered. For tackling this, a deep learning-inspired Facial Emotion Recognition (FER) model that recognizes facial emotions in real-time is proposed in this paper. The proposed Convolutional Neural Network (CNN) model was evaluated on the popular facial emotion recognition dataset of FER2013 using different performance evaluation metrics. Additionally, showcased it as a web application that runs the FER model in real-time to provide audio feedback to the visually impaired to interpret respective emotions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

GLNet: Low-light image enhancement via grayscale priors 13

GLNet: Low-light image enhancement via grayscale priors

引用

13th International Conference on Signal Processing Systems, ICSPS 2021

作者： Guo, Li Xie, Junwei Xue, Yuyang Li, Ru Zheng, Weixin Tong, Tong Gao, Qinquan College of Physics and Information Engineering Fuzhou University Tsukuba Japan Fujian Key Lab of Medical Instrumentation & Pharmaceutical Technology Fuzhou University Tsukuba Japan Imperial Vision Technology Fujian China Department of Computer Science University of Tsukuba Tsukuba Japan

ISBN: (数字)9781510653184

ISBN: (纸本)9781510653177

Low-light images are generally produced by shooting in a low light environment or a tricky shooting angle, which not only affect people's perception, but also leads to the bad performance of some artificial intelligence algorithms, such as object detection, super-resolution, and so on. There are two difficulties in the low-light enhancement algorithm: in the first place, applying image processing algorithms independently to each low-light image often leads to the color distortion;the second is the need to restore the texture of the extremely low-light area. To address these issues, we present two novel and general approaches: firstly, we propose a new loss function to constrain the ratio between the corresponding RGB pixel values on the low-light image and the high-light image;secondly, we propose a new framework named GLNet, which uses the dense residual connection block to obtain the deep features of the low-light images, and design a gray scale channel network branch to guide the texture restoration on the RGB channels by enhancing the grayscale image. The ablation experiments have demonstrated the effectiveness of the proposed module in this paper. Extensive quantitative and perceptual experiments show that our approach obtains state-of-the-art performance on the public dataset. © COPYRIGHT SPIE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

A volumetric change detection framework using UAV oblique photogrammetry–a case study of ultra-high-resolution monitoring of progressive building collapse

引用

International Journal of Digital Earth 2021年第11期14卷 1705-1720页

作者： Ningli Xu Debao Huang Shuang Song Xiao Ling Chris Strasbaugh Alper Yilmaz Halil Sezen Rongjun Qin Geospatial Data Analytics Lab The Ohio State UniversityColumbusOHUSA Photogrammetric Computer Vision Lab The Ohio State UniversityColumbusOHUSA Department of Civil Environmental and Geodetic EngineeringThe Ohio State UniversityColumbusOHUSA Department of Electrical and Computer Engineering The Ohio State UniversityColumbusOHUSA Translational Data Analytics Institute The Ohio State UniversityColumbusOHUSA Engineering Technology Services The Ohio State UniversityColumbusOHUSA

In this paper,we present a case study that performs an unmanned aerial vehicle(UAV)based fine-scale 3D change detection and monitoring of progressive collapse performance of a building during a demolition ***-temporal oblique photogrammetry images are collected with 3D point clouds generated at different stages of the *** geometric accuracy of the generated point clouds has been evaluated against both airborne and terrestrial LiDAR point clouds,achieving an average distance of 12 cm and 16 cm for roof and façade *** propose a hierarchical volumetric change detection framework that unifies multi-temporal UAV images for pose estimation(free of ground control points),reconstruction,and a coarse-to-fine 3D density change *** work has provided a solution capable of addressing change detection on full 3D time-series datasets where dramatic scene content changes are presented *** change detection results on the building demolition event have been evaluated against the manually marked ground-truth changes and have achieved an F-1 score varying from 0.78 to 0.92,with consistently high precision(0.92–0.99).Volumetric changes through the demolition progress are derived from change detection and have been shown to favorably reflect the qualitative and quantitative building demolition progression.

关键词： 3D change detection multitemporal data registration oblique photogrammetry

来源：评论

学校读者我要写书评

暂无评论

Reswave-Net: A wavelet based Residual U-Net for Brain Tumour Segmentation and Overall Survival Prediction

Reswave-Net: A wavelet based Residual U-Net for Brain Tumour...

引用

International Conference on Control Communication & Computing India (ICCC)

作者： Shilna E. Athira K. Vinod Jeena R. S. Anurenjan P. R. Sreeni K. G. Computer Vision Lab College of Engineering Trivandrum Kerala APJ Abdul Kalam Technological University Kerala India Government Engineering College Idukki Kerala Rajiv Gandhi Institute of Technology Kottayam Kerala

A brain tumour is an abnormality in brain tissue that may cause harm to the nervous system and in severe cases can lead to death. Being a disease with a high mortality rate, the detection and accurate segmentation of brain tumour sub-regions is a crucial task in the disease diagnosis and treatment procedure. The manual segmentation process requires anatomical knowledge, is expensive, time-consuming, and inaccurate due to human errors. So, automatic and reliable segmentation methods are required; however, the large spatial and structural variability among brain tumours make automatic segmentation a challenging problem. This work proposes Reswave-Net, a deep learning network using an encoder-decoder (U-Net) architecture with residual connections to automate and standardize the task of tumour segmentation, which also incorporates wavelet decomposition of the input images. The network is trained and evaluated on Brain Tumour Segmentation (BraTS) Challenge-2020 dataset and achieves a mean Dice Score of 87.36%, 70.45%, and 72.55% and the Hausdorff distance of 6.87, 34.16 and 23.42 for the whole tumour, enhancing tumour and tumour core, respectively. For overall survival prediction, a random forest model is used where the radiomic features extracted from the image and age of the subject are used for training. The model achieves an accuracy of 58.4%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

NTIRE 2023 Challenge on Night Photography Rendering

NTIRE 2023 Challenge on Night Photography Rendering

引用

2023 IEEE/CVF Conference on computer vision and Pattern Recognition Workshops, CVPRW 2023

作者： Shutova, Alina Ershov, Egor Perevozchikov, Georgy Ermakov, Ivan Banić, Nikola Timofte, Radu Collins, Richard Efimova, Maria Terekhin, Arseniy Zini, Simone Rota, Claudio Buzzelli, Marco Bianco, Simone Schettini, Raimondo Lei, Chunxia Wang, Tingniao Wang, Song Liu, Shuai Feng, Chaoyu Shao, Guangqi Wang, Hao Wang, Xiaotao Lei, Lei Xu, Lu Zhang, Chao Wang, Yasi Guo, Jin Sun, Yangfan Liu, Tianli Hao, Dejun Kinli, Furkan Özcan, Bariş Kiraç, Furkan Chung, Hyerin Lee, Nakyung Kwak, Sung Keun Conde, Marcos Elezabi, Omar Seizinger, Tim Vasluianu, Florin Hsieh, Chia-Hsuan Chen, Wei-Ting Yang, Hao-Hsiang Huang, Zhi-Kai Chang, Hua-En Chen, I-Hsiang Chen, Yi-Chung Chiang, Yuan-Chun Kuo, Sy-Yen Russia Gideon Brothers Croatia Swiss Federal Institute of Technology in Zurich Switzerland University of Milano - Bicocca Italy Zhejiang Dahua Technology China Xiaomi Inc. China Beijing Samsung Teleconmmunication R&D Center China Vision and Graphics Lab Özyeǧin University Turkey Fishency Innovation Norway T-Fashion Inc. Turkey CJ OliveNetworks AI Research Korea Republic of University of Wuerzburg Computer Vision Lab CAIDAS Germany ServiceNow United States Graduate Institute of Electronics Engineering National Taiwan University Taiwan Department of Electrical Engineering National Taiwan University Taiwan Graduate Institute of Communication Engineering National Taiwan University Taiwan University of Würzburg Germany

ISBN: (纸本)9798350302493

This paper presents a review of the NTIRE 2023 challenge on night photography rendering. The goal of the challenge was to find solutions that process raw camera images taken in nighttime conditions conditions, and thereby produce a photo-quality output images in the standard RGB (sRGB) space. Unlike the previous year's competition, participants were not provided with a large training dataset for the target sensor. Instead, this time they were given images of a color checker illuminated by a known light source. To evaluate the results, a sufficient number of viewers were asked to assess the visual quality of the proposed solutions, considering the subjective nature of the task. The highest ranking solutions were further ranked by Richard Collins, a renowned photographer. The top ranking participants' solutions effectively represent the state-of-the-art in nighttime photography rendering. More results can be found at https://***/ © 2023 IEEE.

关键词： Photography

来源：评论

学校读者我要写书评

暂无评论

OpenKG Chain:A Blockchain Infrastructure for Open Knowledge Graphs

引用

Data Intelligence 2021年第2期3卷 205-227页

作者： Huajun Chen Ning Hu Guilin Qi Haofen Wang Zhen Bi Jie Li Fan Yang College of Computer Science&AZFT Knowledge Engine Lab Zhejiang UniversityHangzhou 310058China Ontology Foundation Venture Drive#11-31Vision ExchangeSingapore 608526Singapore School of Computer Science and Engineering Southeast UniversityNanjing 211189China Intelligent Big Data Visualization Lab Tongji UniversityShanghai 200092China

The early concept of knowledge graph originates from the idea of the semantic Web,which aims at using structured graphs to model the knowledge of the world and record the relationships that exist between *** publishing knowledge bases as open data on the Web has gained significant *** China,Chinese Information Processing Society of China(CIPS)launched the OpenKG in 2015 to foster the development of Chinese Open Knowledge *** existing open knowledge-based programs,OpenKG chain is envisioned as a blockchain-based open knowledge *** article introduces the first attempt at the implementation of sharing knowledge graphs on OpenKG chain,a blockchain-based trust *** have completed the test of the underlying blockchain platform,and the on-chain test of OpenKG’s data set and tool set sharing as well as fine-grained knowledge crowdsourcing at the triple *** have also proposed novel definitions:K-Point and OpenKG Token,which can be considered to be a measurement of knowledge value and user value.1,033 knowledge contributors have been involved in two months of testing on the blockchain,and the cumulative number of on-chain recordings triggered by real knowledge consumers has reached 550,000 with an average daily peak value of more than 10,*** the first time,we have tested and realized on-chain sharing of knowledge at entity/triple granularity *** present,all operations on the data sets and tool sets at ***,as well as the triplets at OpenBase,are recorded on the chain,and corresponding value will also be generated and assigned in a trusted *** this effort,OpenKG chain looks forward to providing a more credible and traceable knowledge-sharing platform for the knowledge graph community.

关键词： Open knowledge graph Blockchain Decentralized distributed network Decentralized knowledge graph

来源：评论

学校读者我要写书评

暂无评论

CLIMS++: Cross Language Image Matching with Automatic Context Discovery for Weakly Supervised Semantic Segmentation

引用

International Journal of computer vision 2025年 1-20页

作者： Xie, Jinheng Deng, Songhe Hou, Xianxu Luo, Zhaochuan Shen, Linlin Huang, Yawen Zheng, Yefeng Shou, Mike Zheng Computer Vision Institute School of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China Department of Computer Science Wenzhou-Kean University Wenzhou325060 China Guangdong Provincial Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen China Show Lab National University of Singapore Singapore639798 Singapore School of AI and Advanced Computing Xi’an Jiaotong-Liverpool University Suzhou215400 China Jarvis Research Center Tencent YouTu Lab Shenzhen518703 China

While promising results have been achieved in weakly-supervised semantic segmentation (WSSS), limited supervision from image-level tags inevitably induces discriminative reliance and spurious relations between target classes and background regions. Thus, Class Activation Map (CAM) usually tends to activate discriminative object regions and falsely includes lots of class-related backgrounds. Without pixel-level supervisions, it could be very difficult to enlarge the foreground activation and suppress those false activation of background regions. In this paper, we propose a novel framework of Cross Language Image Matching with Automatic Context Discovery (CLIMS++), based on the recently introduced Contrastive Language-Image Pre-training (CLIP) model, for WSSS. The core idea of our framework is to introduce natural language supervision to activate more complete object regions and suppress class-related background regions in CAM. In particular, we design object, background region, and text label matching losses to guide the model to excite more reasonable object regions of each category. In addition, we propose to automatically find spurious relations between foreground categories and backgrounds, through which a background suppression loss is designed to suppress the activation of class-related backgrounds. The above designs enable the proposed CLIMS++ to generate a more complete and compact activation map for the target objects. Extensive experiments on PASCAL VOC 2012 and MS COCO 2014 datasets show that our CLIMS++ significantly outperforms the previous state-of-the-art methods. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Online Self-distillation and Self-modeling for 3D Brain Tumor Segmentation

引用

IEEE Journal of Biomedical and Health Informatics 2025年 PP卷 PP页

作者： Pang, Yan Li, Yunhao Huang, Teng Liang, Jiaming Wang, Zhen Dong, Changyu Kuang, Dongyang Hu, Ying Chen, Hao Lei, Tim Wang, Qiong The Guangdong Provincial Key Laboratory of Computer Vision and Virtual Reality Technology Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China The School of Artificial Intelligence Guangzhou University China The Zhejiang Lab Hangzhou China Sun Yat-sen University China The Department of Computer Science and Engineering The Department of Chemical and Biological Engineering Hong Kong University of Science and Technology China The Department of Electrical Engineering University of Colorado Denver United States

In the specialized domain of brain tumor segmentation, supervised segmentation approaches are hindered by the limited availability of high-quality labeled data, a condition arising from data privacy concerns, significant costs, and ethical issues. In response to this challenge, this paper presents a training framework that adeptly integrates a plug-and-play component, MOD, into current supervised learning models, boosting their efficacy in scenarios with limited data. The MOD consists of an Online Tokenizer and a Dense Predictor, which employs self-distillation and self-modeling on masked patches, promoting swift convergence and efficient representation learning. During the inference phase, the plug-and-play MOD component is excluded, preserving the computational efficiency of the original model without incurring extra processing costs. We substantiated the value of our approach through experiments on leading 3D brain tumor segmentation baselines. Remarkably, models augmented with the MOD consistently showcased superior results, achieving elevated Dice coefficients and HD95 scores on two datasets: BraTS 2021 and MSD 2019 Task-01 Brain Tumor. Code: https://***/aigzhusmart/MOD © 2013 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Speech Emotion Recognition, Datasets, Features and Models: A Review

Speech Emotion Recognition, Datasets, Features and Models: A...

引用

International Conference on Control Communication & Computing India (ICCC)

作者： V. Geethu M. K. Vrindha P. R. Anurenjan S. Deepak K. G. Sreeni Computer Vision Lab College of Engineering Trivandrum Kerala India APJ Abdul Kalam Technological University Trivandrum Kerala India Rajiv Gandhi Institute of Technology Kottayam Kerala India

There are several ways in which we can express our emotions, such as through our voice, body language, facial expressions, and eye contact. Speech is a crucial means of conveying our feelings and viewpoints to others, and as a result, emotions play a vital role in human communication and connection. Therefore, we can use speech to understand our own emotions as well as those of others. With the advancement of deep learning, speech emotion recognition (SER) has gained increasing importance in fields like human-machine interaction. SER technology can analyze the emotional content of speech and provide insights into the speaker’s emotional state. This technology has the potential to improve communication and interaction between humans and machines, as well as offer valuable insights in various fields such as psychology, marketing, and customer *** review covers a range of techniques for identifying emotions in speech and speech-text pairs. A variety of deep learning and machine learning techniques have been developed in this field, including CNN-based, LSTM-based, and SVM-based systems for processing voice signals, and BERT, GRU, and other networks for processing text. The main focus of the study is on feature extraction, classification techniques, widely used datasets, and result comparison. In other words, the review discusses how different methods can be used to extract relevant features from speech and text data, how these features can be used to classify emotions, the datasets that are commonly used for evaluating these techniques, and how the results of different methods compare to each other. Overall, this review aims to provide a comprehensive overview of the current state of the field of speech emotion recognition and identify areas for further research.

关键词：

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：