检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Feng, Runyang Gao, Yixing Ma, Xueqing Tse, Tze Ho Elden Chang, Hyung Jin School of Artificial Intelligence Jilin University China Engineering Research Center of Knowledge-Driven Human-Machine Intelligence Ministry of Education China School of Computer Science University of Birmingham United Kingdom

Temporal modeling is crucial for multi-frame human pose estimation. Most existing methods directly employ optical flow or deformable convolution to predict full-spectrum motion fields, which might incur numerous irrelevant cues, such as a nearby person or background. Without further efforts to excavate meaningful motion priors, their results are suboptimal, especially in complicated spatiotemporal interactions. On the other hand, the temporal difference has the ability to encode representative motion information which can potentially be valuable for pose estimation but has not been fully exploited. In this paper, we present a novel multi-frame human pose estimation framework, which employs temporal differences across frames to model dynamic contexts and engages mutual information objectively to facilitate useful motion information disentanglement. To be specific, we design a multi-stage Temporal Difference Encoder that performs incremental cascaded learning conditioned on multi-stage feature difference sequences to derive informative motion representation. We further propose a Representation Disentanglement module from the mutual information perspective, which can grasp discriminative task-relevant motion signals by explicitly defining useful and noisy constituents of the raw motion features and minimizing their mutual information. These place us to rank No.1 in the Crowd Pose Estimation in Complex Events Challenge on benchmark dataset HiEve, and achieve state-of-the-art performance on three benchmarks PoseTrack2017, PoseTrack2018, and PoseTrack21. © 2023, CC BY-NC-ND.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

A3lign-DFER: Pioneering Comprehensive Dynamic Affective Alignment for Dynamic Facial Expression Recognition with CLIP

arXiv

引用

arXiv 2024年

作者： Tao, Zeng Wang, Yan Lin, Junxiong Wang, Haoran Mai, Xinji Yu, Jiawen Tong, Xuan Zhou, Ziheng Yan, Shaoqi Zhao, Qing Han, Liyuan Zhang, Wenqiang Shanghai Engineering Research Center of AI & Robotics Academy for Engineering & Technology Fudan University Shanghai China School of Information Science and Technology Fudan University Shanghai China Institute of Automation Chinese Academy of Sciences Beijing China Engineering Research Center of AI & Robotics Ministry of Education Academy for Engineering & Technology Fudan University Shanghai China Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China

The performance of CLIP in dynamic facial expression recognition (DFER) task doesn’t yield exceptional results as observed in other CLIP-based classification tasks. While CLIP’s primary objective is to achieve alignment between images and text in the feature space, DFER poses challenges due to the abstract nature of text and the dynamic nature of video, making label representation limited and perfect alignment difficult. To address this issue, we have designed A3lign-DFER, which introduces a new DFER labeling paradigm to comprehensively achieve alignment, thus enhancing CLIP’s suitability for the DFER task. Specifically, our A3lign-DFER method is designed with multiple modules that work together to obtain the most suitable expanded-dimensional embeddings for classification and to achieve alignment in three key aspects: affective, dynamic, and bidirectional. We replace the input label text with a learnable Multi-Dimensional Alignment Token (MAT), enabling alignment of text to facial expression video samples in both affective and dynamic dimensions. After CLIP feature extraction, we introduce the Joint Dynamic Alignment Synchronizer (JAS), further facilitating synchronization and alignment in the temporal dimension. Additionally, we implement a Bidirectional Alignment Training Paradigm (BAP) to ensure gradual and steady training of parameters for both modalities. Our insightful and concise A3lign-DFER method achieves state-of-the-art results on multiple DFER datasets, including DFEW, FERV39k, and MAFW. Extensive ablation experiments and visualization studies demonstrate the effectiveness of A3lign-DFER. The code will be available in the future. © 2024, CC BY.

关键词： Alignment

来源：评论

学校读者我要写书评

暂无评论

A Novel Recurrent Neural Net based Dimension Reduction Technique for Effectual Data Visualization

A Novel Recurrent Neural Net based Dimension Reduction Techn...

引用

Communications and Information Technologies (GCCIT), Global Conference on

作者： Monoo John Y. K. Sharma N. V. Balaji Sahil Suri Nishanthi R Samaksh Goyal School of Mangement - PG Management JAIN (Deemed to be University) Bangalore Karnataka Computer Engineering Vishwakarma Institute of Technology Pune Department of Computer Science Karpagam Academy of Higher Education Coimbatore Centre of Research Impact and Outcome Chitkara University Rajpura Punjab India Department of Computer Science and Engineering Prince Shri Venkateshwara Padmavathy Engineering College Chennai Quantum University Research Center Quantum University Roorkee Uttarakhand India

ISBN: (数字)9798350388916

ISBN: (纸本)9798350388923

t-SNE, UMAP, & Tramp have shown excellent visualization performance on various real-world datasets. These approaches may preserve global structure but not local structure, a trade-off that has always plagued them. computer science RNN machine learning is evolving. RNN models can reliably forecast numeric values or nominal classifications, but a general-purpose mechanism for RNN architecture construction has yet to be devised. computer scientists must use trial-and-error to create accurate RNN models. Models often need multiple input parameters. When there are many input factors, it might be hard to determine which ones affect model prediction the most. science calls this the "curse of dimensionality." Problems may be reduced to two dimensions, however. Humans can plan and understand two-dimensional problems. Visualizing a multi-dimensional dataset may reveal which input factors most affect a nominal or numeric outcome. Identifying these factors may help train RNN models faster by utilizing only input variables that impact the result variable. This study explores RNN training methods and uses dimensional reduction to see and interpret complicated data.

关键词： Training Recurrent neural networks Accuracy Computational modeling Data visualization Predictive models Mathematical models Numerical models Reliability Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Multi-Group Fusion Network with Transformer for Camouflaged Object Detection

SSRN

引用

SSRN 2023年

作者： Song, Xiaogang Zhang, Pengfei Shi, Weiwei Lu, Xiaofeng Hei, Xinhong Xi'an University of Technology School of Computer Science and Engineering Xi'An710048 China Engineering Research Center of Human-machine Integration Intelligent Robot Universities of Shaanxi Province Xi'An710048 China

The accurate identification of highly similar foregrounds and backgrounds is a challenging task in the detection of camouflage objects. Although existing methods improve detection performance by using boundary and texture clues, they are still inadequate for identifying camouflaged objects from the background in complex scenarios. To address this issue, we introduce a novel multi-group fusion network with Transformer, called MGNet, for camouflaged object detection. Specifically, the Transformer architecture is adopted as a backbone network extraction feature to capture long-range dependencies and build global content-aware interactions. Subsequently, a group-fusion attention module is constructed to enhance the discrimination ability of the network to the foreground and background in complex scenes by mining the weight relationship between each channel. Finally, a multi-scale feature fusion module is designed, which adopts a coarse-to-fine fusion strategy to generate high-resolution prediction maps by integrating features step by step, while preserving the global awareness of the network to the maximum extent. Extensive experiments were conducted on four challenging benchmark datasets, which demonstrate that MGNet significantly outperforms the existing 18 state-of-the-art methods under four widely-used evaluation metrics. © 2023, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-based Image Text Processing Research*

Deep Learning-based Image Text Processing Research*

引用

IEEE International Conference on Big Data Security on Cloud (BigDataSecurity), High Performance and Smart Computing (HPSC) and Intelligent Data and Security (IDS)

作者： Huixuan Xiong Kai Jin Jingnian Liu Jiahong Cai Lijun Xiao School of Computer Science and Engineering Hunan University of Science and Technology Hunan Key Laboratory for Service Computing and Novel Software Technology Xiangtan China the College of Computer Science and Electronics Engineering Hunan University Changsha China Big Data Development and Research Center Guangzhou College of Technology and Business Guangzhou China

Deep learning is a powerful multi-layer architecture that has important applications in image processing and text classification. This paper first introduces the development of deep learning and two important algorithms of deep learning: convolutional neural networks and recurrent neural networks. The paper then introduces three applications of deep learning for image recognition, image detection, and image forensics, as well as three text classification methods based on convolutional neural networks, recurrent neural networks, and other text classification methods. Finally, the development trend of deep learning in the field of text and image processing and the difficulties to be further researched are summarised and prospected.

关键词： Deep learning Image forensics Recurrent neural networks Computational modeling Text categorization Feature extraction Data models

来源：评论

学校读者我要写书评

暂无评论

WebCode2M: A Real-World Dataset for Code Generation from Webpage Designs 25

WebCode2M: A Real-World Dataset for Code Generation from Web...

引用

34th ACM Web Conference, WWW 2025

作者： Gui, Yi Li, Zhen Wan, Yao Shi, Yemin Zhang, Hongyu Su, Yi Chen, Bohua Chen, Dongping Wu, Siyuan Zhou, Xing Jiang, Wenbin Jin, Hai Zhang, Xiangliang Huazhong University of Science and Technology Wuhan China Peking University Beijing China Chongqing University Chongqing China Hubei University of Automotive Technology Shiyan China Rabbitpre AI Shenzhen China University of Notre Dame Notre Dame United States National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster Grid Computing Lab School of Computer Science and Technology China

ISBN: (纸本)9798400712746

Automatically generating webpage code from webpage designs can significantly reduce the workload of front-end developers, and recent Multimodal Large Language Models (MLLMs) have shown promising potential in this area. However, our investigation reveals that most existing MLLMs are constrained by the absence of high-quality, large-scale, real-world datasets, resulting in inadequate performance in automated webpage code generation. To fill this gap, this paper introduces WebCode2M, a new dataset comprising 2.56 million instances, each containing a design image along with the corresponding webpage code and layout details. Sourced from real-world web resources, WebCode2M offers a rich and valuable dataset for webpage code generation across a variety of applications. The dataset quality is ensured by a scoring model that filters out instances with aesthetic deficiencies or other incomplete elements. To validate the effectiveness of WebCode2M, we introduce a baseline model based on the Vision Transformer (ViT), named WebCoder, and establish a benchmark for fair comparison. Additionally, we introduce a new metric, TreeBLEU, to measure the structural hierarchy recall. The benchmarking results demonstrate that our dataset significantly improves the ability of MLLMs to generate code from webpage designs, confirming its effectiveness and usability for future applications in front-end design tools. Finally, we highlight several practical challenges introduced by our dataset, calling for further research. © 2025 Copyright held by the owner/author(s).

关键词： Web Design

来源：评论

学校读者我要写书评

暂无评论

Perceptual Audio Object Coding Using Adaptive Subband Grouping with CNN and Residual Block

Perceptual Audio Object Coding Using Adaptive Subband Groupi...

引用

IEEE International Conference on Multimedia and Expo (ICME)

作者： Yulin Wu Ruimin Hu Xiaochen Wang National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University Wuhan China Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University Wuhan China

Spatial audio content is becoming increasingly popular and is regarded as a set of object signals with associated metadata. The object-based content representation is independent of loudspeaker layouts and provides high spatial resolution when reproduced on more loudspeakers. The audio quality of the traditional spatial audio object coding (SAOC) method has severe aliasing distortion, which impairs the immersive listening experience. In this study, we reduce aliasing distortion by perceptual adaptive subband grouping strategy and use the convolutional neural network (CNN) and residual block to build the side information compressing model. Both objective and subjective experiments on benchmark datasets with different bitrates show that the proposed method achieves favorable performance against state-of-the-art methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Temporal Shift for Personality Recognition with Pre-Trained Representations

Temporal Shift for Personality Recognition with Pre-Trained ...

引用

International Symposium on Chinese Spoken Language Processing

作者： Hanzhe Xu Xuefei Liu Cong Cai Kang Zhu Jizhou Cui Ruibo Fu Heng Xie Jianhua Tao Zhengqi Wen Ziping Zhao Guanjun Li Le Wang Hao Lin College of Computer and Information Engineering Tianjin Normal University Institute of Automation Chinese Academy of Sciences Beijing Institute of Technology School of Computer Science and Technology Anhui University School of Information Science and Technology ShanghaiTech University Department of Automation Tsinghua University Beijing National Research Center for Information Science and Technology Tsinghua University Migu Meland Co. Ltd.

ISBN: (数字)9798331516826

ISBN: (纸本)9798331516833

Personality, as a reflection of our inner psychological traits, profoundly influences our behavioral patterns, decision-making processes, and the way we interact with the world. Therefore, personality recognition has gradually become an important research direction in the fields of psychology and human-computer interaction. However, most previous works have focused on using pre-trained models to construct more complex models for feature extraction, neglecting the integration of multi-channel information and the fusion between channel information. To address the issue of multi-channel information integration, we propose and validate an innovative method based on multi-scale convolutional feature extraction and a temporal shift module for personality recognition in the speech modality. We use the wav2vec pre-trained model to extract features and enhance information fusion between channels by introducing multi-scale convolutional kernels and a temporal shift module. It expands the model's perception range and sensitivity to personality. Experimental results show that this method performs well in terms of accuracy (ACC) and concordance correlation coefficient (CCC).

关键词： Human computer interaction Sensitivity Psychology Speech recognition Feature extraction Transformers Reflection Speech processing Kernel Long short term memory

来源：评论

学校读者我要写书评

暂无评论

Query-centered Temporal Community Search via Time-Constrained Personalized PageRank

arXiv

引用

arXiv 2023年

作者： Lin, Longlong Yuan, Pingpeng Li, Rong-Hua Zhu, Chunxue Qin, Hongchao Jin, Hai Jia, Tao College of Computer and Information Science Southwest University Chongqing400715 China National Engineering Research Center for Big Data Technology and System Service Computing Technology and System Laboratory Cluster and Grid Computing Laboratory School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China School of Computer Science and Technololgy Beijing Institute of Technology Beijing100081 China

Existing temporal community search suffers from two defects: (i) they ignore the temporal proximity between the query vertex q and other vertices but simply require the result to include q. Thus, they find many temporal irrelevant vertices (these vertices are called query-drifted vertices) to q for satisfying their cohesiveness, resulting in q being marginalized;(ii) their methods are NP-hard, incurring high costs for exact solutions or compromised qualities for approximate/heuristic algorithms. Inspired by these, we propose a novel problem named query-centered temporal community search to circumvent query-drifted vertices. Specifically, we first present a novel concept of Time-Constrained Personalized PageRank to characterize the temporal proximity between q and other vertices. Then, we introduce a model called β-temporal proximity core, which can combine temporal proximity and structural cohesiveness. Subsequently, our problem is formulated as an optimization task that finds a β-temporal proximity core with the largest β. To solve our problem, we first devise an exact and near-linear time greedy removing algorithm that iteratively removes unpromising vertices. To improve efficiency, we then design an approximate two-stage local search algorithm with bound-based pruning techniques. Finally, extensive experiments on eight real-life datasets and nine competitors show the superiority of the proposed solutions. © 2023, CC BY.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

Multi-View Graph Representation Learning Beyond Homophily

arXiv

引用

arXiv 2023年

作者： Lin, Bei Li, You Gui, Ning Xu, Zhuopeng Yu, Zhiwu School of Computer Science & Engineering Central South University YueLu Street Hunan Changsha410083 China National Engineering Research Center of High-speed Railway Construction Technology Changsha China

Unsupervised graph representation learning(GRL) aims to distill diverse graph information into task-agnostic embeddings without label supervision. Due to a lack of support from labels, recent representation learning methods usually adopt self-supervised learning, and embeddings are learned by solving a handcrafted auxiliary task(so-called pretext task). However, partially due to the irregular non-Euclidean data in graphs, the pretext tasks are generally designed under homophily assumptions and cornered in the low-frequency signals, which results in significant loss of other signals, especially high-frequency signals widespread in graphs with heterophily. Motivated by this limitation, we propose a multi-view perspective and the usage of diverse pretext tasks to capture different signals in graphs into embeddings. A novel framework, denoted as Multi-view Graph Encoder(MVGE), is proposed, and a set of key designs are identified. More specifically, a set of new pretext tasks are designed to encode different types of signals, and a straightforward operation is propxwosed to maintain both the commodity and personalization in both the attribute and the structural levels. Extensive experiments on synthetic and real-world network datasets show that the node representations learned with MVGE achieve significant performance improvements in three different downstream tasks, especially on graphs with heterophily. Source code is available at https://***/G-AILab/MVGE. © 2023, CC BY.

关键词： Embeddings

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：