检索结果-内蒙古大学图书馆

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Yuanhang Zhang Shuang Yang Shiguang Shan Xilin Chen Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS) Institute of Computing Technology CAS Beijing China University of Chinese Academy of Sciences Beijing China

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

We propose a novel strategy, ES 3 , for self-supervised learning of robust audio-visual speech representations from unlabeled talking face videos. While many recent approaches for this task primarily rely on guiding the learning process using the audio modality alone to capture information shared between audio and video, we reframe the problem as the acquisition of shared, unique (modality-specific) and synergistic speech information to address the inherent asymmetry between the modalities. Based on this formulation, we propose a novel “evolving” strategy that progressively builds joint audio-visual speech representations that are strong for both uni-modal (audio & visual) and bi-modal (audio-visual) speech. First, we leverage the more easily learnable audio modality to initialize audio and visual representations by capturing audio-unique and shared speech information. Next, we incorporate video-unique speech information and bootstrap the audio-visual repre-sentations on top of the previously acquired shared knowledge. Finally, we maximize the total audio-visual speech information, including synergistic information to obtain robust and comprehensive representations. We implement ES 3 as a simple Siamese framework and experiments on both English benchmarks and a newly contributed large-scale Mandarin dataset show its effectiveness. In particular, on LRS2-BBC, our smallest model is on par with SoTA models with only 1/2 parameters and 1/8 unlabeled data (223h).

关键词： Visualization Computer vision Self-supervised learning Manuals Benchmark testing Data models Pattern recognition

来源：评论

学校读者我要写书评

暂无评论

An Empirical Study on Neural Networks Pruning: Trimming for Reducing Memory or Workload

An Empirical Study on Neural Networks Pruning: Trimming for ...

引用

Data-driven Optimization of Complex Systems (DOCS), International Conference on

作者： Kaiwen Xiao Xiaodong Cai Ke Xu School of Artificial Intelligence Anhui University Hefei China Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Hefei China

Most of existing studies on neural network pruning only consider memory-based pruning strategies. However pruning for computational workload is often more important in hardware deployments due to a greater focus on model computation reductions. In addition, most pruning schemes restore model accuracy during pruning at the expense of adding hyperparameters, extending training time and training complexity. This work proposes a statistical-based globally soft iterative pruning scheme. With little extra calculation, an extremely sparse model can be obtained without additional hyperparameters and extended training time. Moreover, this work proposes the concept of computational intensity to balance model memory and computational workload during pruning. Focusing on memory orientated pruning, we can achieve $\mathbf{303}\times, \mathbf{100}\times$ and $\mathbf{25}\times$ parameter compression on LeNet-5 (MNIST), VGG (CIFAR-10) and AlexNet (ImageNet) models, respectively. In particular, combined with cluster quantization, the LeNet-5 model parameters can be compressed to $\mathbf{3232}\times$ . Focusing on workload orientated pruning, we can reduce the computation by $\mathbf{7}.\mathbf{6}\times$ on the AlexNet model, without accuracy loss, significantly higher than prior work. In addition, in order to verify the versatility of the pruning method, we also migrate the pruning task to the object detection and implement $\mathbf{10}\times$ parameter compression and $\mathbf{2}.\mathbf{8}\times$ computation compression for YOLOv2 with reduced mAP within 1%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Synthesizing Knowledge-enhanced Features for Real-world Zero-shot Food Detection

arXiv

引用

arXiv 2024年

作者： Zhou, Pengfei Min, Weiqing Song, Jiajun Zhang, Yang Jiang, Shuqiang The Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China The University of Chinese Academy of Sciences Beijing100049 China The Institute of Intelligent Computing Technology Chinese Academy of Sciences Suzhou215124 China

Food computing brings various perspectives to computer vision like vision-based food analysis for nutrition and health. As a fundamental task in food computing, food detection needs Zero-Shot Detection (ZSD) on novel unseen food objects to support real-world scenarios, such as intelligent kitchens and smart restaurants. Therefore, we first benchmark the task of Zero-Shot Food Detection (ZSFD) by introducing FOWA dataset with rich attribute annotations. Unlike ZSD, fine-grained problems in ZSFD like inter-class similarity make synthesized features inseparable. The complexity of food semantic attributes further makes it more difficult for current ZSD methods to distinguish various food categories. To address these problems, we propose a novel framework ZSFDet to tackle fine-grained problems by exploiting the interaction between complex attributes. Specifically, we model the correlation between food categories and attributes in ZSFDet by multi-source graphs to provide prior knowledge for distinguishing fine-grained features. Within ZSFDet, Knowledge-Enhanced Feature Synthesizer (KEFS) learns knowledge representation from multiple sources (e.g., ingredients correlation from knowledge graph) via the multi-source graph fusion. Conditioned on the fusion of semantic knowledge representation, the region feature diffusion model in KEFS can generate fine-grained features for training the effective zero-shot detector. Extensive evaluations demonstrate the superior performance of our method ZSFDet on FOWA and the widely-used food dataset UECFOOD-256, with significant improvements by 1.8% and 3.7% ZSD mAP compared with the strong baseline RRFS. Further experiments on PASCAL VOC and MS COCO prove that enhancement of the semantic knowledge can also improve the performance on general ZSD. Code and dataset are available at https://***/LanceZPF/KEFS. Copyright © 2024, 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

Research on the Structure and Realization of Mixture-of-Experts

Research on the Structure and Realization of Mixture-of-Expe...

引用

International Conference on Computer Engineering and Applications (ICCEA)

作者： Qidong Yan Yingjie Li Ning Ma Fucheng Wan Key Laboratory of Linguistic and Cultural Computing Ministry of Education Northwest Minzu University Lanzhou China Key Laboratory of China's Ethnic Languages and Intelligent Processing of Gansu Province Northwest Minzu University Lanzhou China

ISBN: (数字)9798350386776

ISBN: (纸本)9798350386783

The development of Large Language Model (LLM) has reached a bottleneck, and in order to solve these problems it is necessary to continue to increase the complexity of the model. The parameters of LLM become larger and larger with the practical requirements of different application scenarios, which increases the difficulty of training as well as inference. One of the approaches to deal with the question is also known as Mixture-of-Experts (MoE), which is based on gated networks. The Mixture-of-Experts is a sparse gate-controlled deep learning model that mainly consists of a set of expert models and a gated model. The basic idea of the Mixture-of-Experts is to partition the input data into multiple regions based on the task type and assign one or more expert models to the data in each region. It is able to achieve excellent performance in complex tasks and flexibly adapt to different input distributions and task scenarios. Its high sparsity improves computational efficiency, and its specialized design enhances the ability to model complex data structures.

关键词： Training Deep learning Adaptation models Computational modeling Large language models Logic gates Data structures

来源：评论

学校读者我要写书评

暂无评论

Multi-View Consistent 3D GAN Inversion via Bidirectional Encoder

Multi-View Consistent 3D GAN Inversion via Bidirectional Enc...

引用

International Conference on Automatic Face and Gesture Recognition

作者： Haozhan Wu Hu Han Shiguang Shan Xilin Chen Key Laboratory of Intelligent Information Processing Chinese Academy of Sciences (CAS) Institute of Computing Technology CAS Beijing China University of the Chinese Academy of Sciences Beijing China

ISBN: (数字)9798350394948

ISBN: (纸本)9798350394955

3D GAN inversion enables not only 3D reconstruction from a 2D image, but also novel view synthesis and image editing. Existing works ensure the novel view synthesis quality by constraining the synthesized views to conform to the real image distribution. However, most of the methods did not consider the multi-view consistency, i.e., different photos of the same 3D scene via 3D GAN inversion can be inverted to the same 3D scene. In this paper, we propose a bidirectional encoder (BiDiE) for 3D GAN inversion that can improve the multi-view consistency and alleviate the interference of camera parameter prediction errors. On the one hand, the bidirectional encoder takes real images as input, estimates the camera parameters, and performs 3D reconstruction. On the other hand, the bidirectional encoder takes randomly sampled latent code and camera parameters as input, and generates synthesized images to assist in the latent code learning process. In addition, we extend the latent space from W+ to W++ to improve its reconstruction and editing capabilities. Experiments on the FFHQ, CelebA-HQ and Multi-PIE datasets prove that our proposed method outperforms state-of-the-art methods in multi-view consistent reconstruction as well as editing capability. 1 1 Code and datasets are available at https://***/WHZMM/BiDiE

关键词： Three-dimensional displays Codes Face recognition Interference Gesture recognition Cameras Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

RALT—A film review sentiment classification model

RALT—A film review sentiment classification model

引用

International Conference on Computer Engineering and Applications (ICCEA)

作者： Ning Chen Xiangzhen He Xianghe Meng Yugang Dai Key Laboratory of Linguistic and Cultural Computing Ministry of Education (Northwest Minzu University) Lanzhou China Key Laboratory of Minzu Languages and Cultures Intelligent Information Processing Gansu Province (Northwest Minzu University) Lanzhou China

ISBN: (数字)9798350386776

ISBN: (纸本)9798350386783

With the development of the internet and film industry, filmgoers have published a large amount of review texts in many areas such as film analysis, film sharing, and social media. The interpretation of review texts is diverse and strongly environmentally dependent, which adds more difficulty to the task of text classification. How to quickly and accurately judge the sentiment tendency of review texts and classify them accurately, and improve the application value of review text data has become one of the popular research directions in natural language processing (NLP). This article proposes a film review sentiment analysis model RALT, which combines a text convolutional neural network (TextCNN) with a self-attention mechanism and a fused long short-term memory (LSTM). The model uses the RoBERTa Employing a pre-trained model to produce dynamic textual word embeddings addresses the intricacies associated with polysemy. Sentiment classification experiments were conducted on the IMDB dataset. The experiments were conducted by comparing BERT and RoBERTa with TextCNN, RNN, LSTM, BI-LSTM and other models, and it was found that the proposed model achieved the best accuracy and F1-score in the classification *** experiment showed that the model is suitable for the field of film review sentiment classification.

关键词： Accuracy Reviews Social networking (online) 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

A Gong-Si-Shaped Heterogeneous Circular-Polarization Array for 5G Application 5

A Gong-Si-Shaped Heterogeneous Circular-Polarization Array f...

引用

5th IEEE International Conference on Electronic Information and Communication Technology, ICEICT 2022

作者： Zhou, Wei Qi, Junwei Li, Yingsong Huang, Zhixiang Yang, Lixia College of Information and Communication Engineering Harbin Engineering University Harbin150001 China Anhui Province Key Laboratory of Target Recognition and Feature Extraction LuAn230088 China Anhui University Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Hefei China

ISBN: (数字)9781665472111

ISBN: (纸本)9781665472111

In this work, a Gong-Si-shaped circularly polarized (CP) array with power driver for 5G is proposed. The radiator of the constructed CP array consists of two Chinese-like characters ' and '' with the similar size and pronunciation of 'gong' and 'si', Two L-shaped feeding probes of the same size are connected to the designed power driver printed on the bottom layer of ground plate, and the array can radiate circular polarization electromagnetic waves by changing the positions of the L-shaped probes and the angle of the horizontal parts of the L-shaped probes. From the simulating and optimizing, it is finally obtained that the -15 dB bandwidth of modeled Gong-Si-like CP antenna lies in 2.49 - 2.81 GHz and the presented Gong-Si-like CP antenna has a axial bandwidth covering 2.36 - 2.82 GHz, which meets the n41 5G band implementing at 2.515 - 2.675 GHz. © 2022 IEEE.

关键词： Circular polarization

来源：评论

学校读者我要写书评

暂无评论

Towards Robust Semantic Segmentation against Patch-based Attack via Attention Refinement

arXiv

引用

arXiv 2024年

作者： Yuan, Zheng Zhang, Jie Wang, Yude Shan, Shiguang Chen, Xilin Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China

The attention mechanism has been proven effective on various visual tasks in recent years. In the semantic segmentation task, the attention mechanism is applied in various methods, including the case of both Convolution Neural Networks (CNN) and Vision Transformer (ViT) as backbones. However, we observe that the attention mechanism is vulnerable to patch-based adversarial attacks. Through the analysis of the effective receptive field, we attribute it to the fact that the wide receptive field brought by global attention may lead to the spread of the adversarial patch. To address this issue, in this paper, we propose a Robust Attention Mechanism (RAM) to improve the robustness of the semantic segmentation model, which can notably relieve the vulnerability against patch-based attacks. Compared to the vallina attention mechanism, RAM introduces two novel modules called Max Attention Suppression and Random Attention Dropout, both of which aim to refine the attention matrix and limit the influence of a single adversarial patch on the semantic segmentation results of other positions. Extensive experiments demonstrate the effectiveness of our RAM to improve the robustness of semantic segmentation models against various patch-based attack methods under different attack settings. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Two Physics-Based Schwarz Preconditioners for Three-Temperature Radiation Diffusion Equations in High Dimensions

引用

Communications in Computational Physics 2022年第8期32卷 829-849页

作者： Xiaoqiang Yue Jianmeng He Xiaowen Xu Shi Shu Libo Wang National Center for Applied Mathematics in Hunan Key Laboratory of Intelligent Computing&Information Processing of Ministry of EducationHunan Key Laboratory for Computation and Simulation in Science and EngineeringXiangtan UniversityXiangtan 411105P.R.China Laboratory of Computational Physics Institute of Applied Physics and ComputationalMathematicsBeijing 100094P.R.China College of Information Science and Technology JinanUniversityGuangzhou 510632P.R.China

We concentrate on the parallel,fully coupled and fully implicit solution of the sequence of 3-by-3 block-structured linear systems arising from the symmetrypreserving finite volume element discretization of the unsteady three-temperature radiation diffusion equations in high *** this article,motivated by[***,***,***,SIAM *** ***.33(2012)653–680]and[***,***,***,***.442(2021)110513],we aim to develop the additive and multiplicative Schwarz preconditioners subdividing the physical quantities rather than the underlying domain,and consider their sequential and parallel implementations using a simplified explicit decoupling factor approximation and algebraic multigrid subsolves to address such linear ***,computational efficiencies and parallel scalabilities of the proposed approaches are numerically tested in a number of representative real-world capsule implosion benchmarks.

关键词： Radiation diffusion equations Schwarz methods algebraic multigrid parallel and distributed computing

来源：评论

学校读者我要写书评

暂无评论

Evolutionary Neural Architecture Search for Transformer in Knowledge Tracing

arXiv

引用

arXiv 2023年

作者： Yang, Shangshang Yu, Xiaoshan Tian, Ye Yan, Xueming Ma, Haiping Zhang, Xingyi Anhui University China Guangdong University of Foreign Studies China Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education China

Transformer has achieved excellent performance in the knowledge tracing (KT) task, but they are criticized for the manually selected input features for fusion and the defect of single global context modelling to directly capture students’ forgetting behavior in KT, when the related records are distant from the current record in terms of time. To address the issues, this paper first considers adding convolution operations to the Transformer to enhance its local context modelling ability used for students’ forgetting behavior, then proposes an evolutionary neural architecture search approach to automate the input feature selection and automatically determine where to apply which operation for achieving the balancing of the local/global context modelling. In the search space design, the original global path containing the attention module in Transformer is replaced with the sum of a global path and a local path that could contain different convolutions, and the selection of input features is also considered. To search the best architecture, we employ an effective evolutionary algorithm to explore the search space and also suggest a search space reduction strategy to accelerate the convergence of the algorithm. Experimental results on the two largest and most challenging education datasets demonstrate the effectiveness of the architecture found by the proposed approach. Copyright © 2023, The Authors. All rights reserved.

关键词： Convolution

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：