检索结果-内蒙古大学图书馆

2022 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2022

作者： Zhao, Zihao Jiang, Min Guo, Jia Yang, Xiaoyu Hu, Yudie Zhou, Xianlong Wuhan University of Science and Technology College of Computer Science and Technology Wuhan430065 China Wuhan University of Science and Technology Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System Wuhan430065 China

ISBN: (数字)9781665452588

ISBN: (纸本)9781665452588

Raindrop removal for in-vehicle camera images is useful for surveillance and analysis system such as intelligent driving or case diagnosis. As in rainy days, images taken by in-vehicle cameras such as monitor or drive recorder often suffer from noticeable image degradation, and make it difficult to identify the objects on the road. With the uncertainty of raindrop distribution and the complexity of raindrop status, the attached raindrops of in-vehicle camera images can have different effects. And the large-diameter raindrop will greatly reduce the image quality. However, most of the popular deraining algorithms have been successful in recovering images with small raindrop noise and tiny image distortion, but fails to restore those with raindrops covering large areas. We propose a single image raindrop removal network based on Generative Adversarial Network for in-vehicle camera images. We incorporate two overlapping attention layers into the deraining network, which adopt task-driven visual attention and content perception mechanism. The former obtains features by recursive neural network to guide network more interested in the raindrop regions and the surrounding structures, while the content perception schema extracts the features far from raindrops. Moreover, we collect a new benchmark for In-Vehicle Camera Image Deraining (IVCRID), it chooses in-vehicle camera images from various deraining dataset. The experimental results show that the proposed deraining network outperforms other state-of-the-art raindrop removal methods in image recovering test on the IVCRID dataset and also has the best performance on the experiments of traffic objects detection. © 2022 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Attention Residual Network for White Blood Cell Classification with WGAN Data Augmentation 11

Attention Residual Network for White Blood Cell Classificati...

引用

11th International Conference on information technology in Medicine and Education, ITME 2021

作者： Zhao, Meng Jin, Lingmin Teng, Shenghua Li, Zuoyong College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou China

ISBN: (纸本)9781665406796

In medicine, white blood cell (WBC) classification plays an important role in clinical diagnosis and treatment. Due to the similarity between classes and lack of training data, the precise classification of WBC is still challenging. To alleviate this problem, we propose an attention residual network for WBC image classification on the basis of data augmentation. Specifically, the attention residual network is composed of multiple attention residual blocks, an adaptive average pooling layer, and a full connection layer. The channel attention mechanism is introduced in each residual block to use the feature maps of WBC learned by a high layer to generate the attention map for a low layer. Each attention residual block also introduces depth separable convolution to extract the feature of WBC and decrease the training costs. The Wasserstein Generative adversarial network (WGAN) is used to create synthetic instances to enhance the size of training data. Experiments on two image datasets show the superiority of the proposed method over several state-of-the-art methods. © 2021 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

PRTGS: Precomputed Radiance Transfer of Gaussian Splats for Real-Time High-Quality Relighting

arXiv

引用

arXiv 2024年

作者： Guo, Yijia Bai, Yuanxi Hu, Liwen Guo, Ziyi Liu, Mianzhi Cai, Yu Huang, Tiejun Ma, Lei State Key Laboratory of Multimedia Information Processing School of Computer Science Peking University Beijing China College of Future Technology Peking University Beijing China

We proposed Precomputed Radiance Transfer of Gaussian Splats (PRTGS), a real-time high-quality relighting method for Gaussian splats in low-frequency lighting environments that captures soft shadows and interreflections by precomputing 3D Gaussian splats' radiance transfer. Existing studies have demonstrated that 3D Gaussian splatting (3DGS) outperforms neural fields' efficiency for dynamic lighting scenarios. However, the current relighting method based on 3DGS is still struggling to compute high-quality shadow and indirect illumination in real time for dynamic light, leading to unrealistic rendering results. We solve this problem by precomputing the expensive transport simulations required for complex transfer functions like shadowing, the resulting transfer functions are represented as dense sets of vectors or matrices for every Gaussian splat. We introduce distinct precomputing methods tailored for training and rendering stages, along with unique ray tracing and indirect lighting precomputation techniques for 3D Gaussian splats to accelerate training speed and compute accurate indirect lighting related to environment light. Experimental analyses demonstrate that our approach achieves state-of-the-art visual quality while maintaining competitive training times and allows high-quality real-time (30+ fps) relighting for dynamic light and relatively complex scenes at 1080p resolution. Copyright © 2024, The Authors. All rights reserved.

关键词： Ray tracing

来源：评论

学校读者我要写书评

暂无评论

Training-free Diffusion Model Adaptation for Variable-Sized Text-to-Image Synthesis

arXiv

引用

arXiv 2023年

作者： Jin, Zhiyu Shen, Xuli Li, Bin Xue, Xiangyang Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China UniDT Technology United States

Diffusion models (DMs) have recently gained attention with state-of-the-art performance in text-to-image synthesis. Abiding by the tradition in deep learning, DMs are trained and evaluated on the images with fixed sizes. However, users are demanding for various images with specific sizes and various aspect ratio. This paper focuses on adapting text-to-image diffusion models to handle such variety while maintaining visual fidelity. First we observe that, during the synthesis, lower resolution images suffer from incomplete object portrayal, while higher resolution images exhibit repetitively disordered presentation. Next, we establish a statistical relationship indicating that attention entropy changes with token quantity, suggesting that models aggregate spatial information in proportion to image resolution. The subsequent interpretation on our observations is that objects are incompletely depicted due to limited spatial information for low resolutions, while repetitively disorganized presentation arises from redundant spatial information for high resolutions. From this perspective, we propose a scaling factor to alleviate the change of attention entropy and mitigate the defective pattern observed. Extensive experimental results validate the efficacy of the proposed scaling factor, enabling models to achieve better visual effects, image quality, and text alignment. Notably, these improvements are achieved without additional training or fine-tuning techniques. Copyright © 2023, The Authors. All rights reserved.

关键词： Aspect ratio

来源：评论

学校读者我要写书评

暂无评论

Research on Collaborative Filtering Algorithm of E-commerce Based on Flink

Research on Collaborative Filtering Algorithm of E-commerce ...

引用

International Conference on Intelligent Control and information processing (ICICIP)

作者： Haohao Zheng Xiaonan Luo Xiaoshu Zhu School of Computer Science and Information Security Guilin University of Electronic Technology Guilin China Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin University of Electronic Technology Guilin China

ISBN: (数字)9798331516147

ISBN: (纸本)9798331516154

In online shopping, a person’s interest in a product is closely related to whether they will purchase it Analyzing people’s interest in various products on time, along with product recommendations, can increase purchase rates. However, in many cases, We only know user implicit feedback, such as browsing or favoriting items, and cannot directly access user explicit feedback, such as product ratings. Moreover, users have limited exposure to products, resulting in a sparse user-item matrix. The traditional Alternating Least Squares (ALS) algorithm can partially address the is of of matrix sparsity but its accuracy in recommending products based on implicit feedback alone is insufficient. This paper proposes an improved ALS algorithm that converts implicit feedback into explicit ratings. We calculate the browsing frequency of a product based on user browsing counts. For a given product, we represent user preference by suimming the browsing frequencies of all products with browsing frequencies less than or equal to that of the product. Then, by multiplying the preference level by the product’s maximum rating, we derive a product rating, achieving a transformation from implicit feedback to explicit scores. Additionally, traditional collaborative filtering algorithms often use offline data, which may not align with user current preferences. To improve the algorithm’s accuracy, we also utilize Flink for real-time data processing, allowing us to capture the most recent feedback data. This study tests the approach using the UserBehavior dataset from the publicly available Tianchi dataset. Experimental results indicate improvements in both Root Mean Square Error (RMSE) and Mean Absolute Error (MAE).

关键词： Accuracy Collaborative filtering Prediction algorithms Real-time systems Electronic commerce Sparse matrices User preference Matrix converters Root mean square Intelligent control

来源：评论

学校读者我要写书评

暂无评论

Underwater Image Restoration Based on Improved SwinIR Model

Underwater Image Restoration Based on Improved SwinIR Model

引用

International Conference on Intelligent Control and information processing (ICICIP)

作者： Yanke Lu Fang Li Xiaonan Luo Guangxi Key Laboratory of Image and and Graphic Intelligent Processing Guilin University of Electronic Technology Guilin China School of Computer Science and Information Security Guilin University of Electronic Technology Guilin China Center for Applied Mathematics of Guangxi (GUET) Guilin China Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin China Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin University of Electronic Technology Guilin China

ISBN: (数字)9798331516147

ISBN: (纸本)9798331516154

Image restoration is a classic foundational visual task, aimed at recovering damaged images, such as those affected by compression, blurring, or noise, to high-definition clarity. Although current image enhancement techniques largely rely on convolutional neural networks, the efficient capabilities of Transformers demonstrated in complex visual tasks have not yet been extensively applied in the field of image restoration. This paper presents an optimized image restoration strategy based on the SwinlR model, which is applied to the task of underwater image restoration. In the shallow feature extraction phase, deep separable convolution technology is employed; while in the deep feature extraction phase, a Spatial Gated Feedforward Network (SGFN) is utilized. The experimental data suggests that the novel approach surpasses conventional techniques in effectiveness and has effectively minimized both the model's parameter quantity and computational demands.

关键词： Visualization Convolution Computational modeling Noise information processing Logic gates Feature extraction Transformers Image restoration Intelligent control

来源：评论

学校读者我要写书评

暂无评论

Mitigating Low-Frequency Bias: Feature Recalibration and Frequency Attention Regularization for Adversarial Robustness

arXiv

引用

arXiv 2024年

作者： Zhang, Kejia Weng, Juanjuan Cai, Yuanzheng Luo, Zhiming Li, Shaozi Department of Artificial Intelligence Xiamen University Xiamen361005 China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou350108 China College of Information Science and Technology Jinan University Guangzhou510632 China

Ensuring the robustness of deep neural networks against adversarial attacks remains a fundamental challenge in computer vision. While adversarial training (AT) has emerged as a promising defense strategy, our analysis reveals a critical limitation: AT-trained models exhibit a bias toward low-frequency features while neglecting high-frequency components. This bias is particularly concerning as each frequency component carries distinct and crucial information: low-frequency features encode fundamental structural patterns, while high-frequency features capture intricate details and textures. To address this limitation, we propose High-Frequency Feature Disentanglement and Recalibration (HFDR), a novel module that strategically separates and recalibrates frequency-specific features to capture latent semantic cues. We further introduce frequency attention regularization to harmonize feature extraction across the frequency spectrum and mitigate the inherent low-frequency bias of AT. Extensive experiments demonstrate our method’s superior performance against white-box attacks and transfer attacks, while exhibiting strong generalization capabilities across diverse scenarios. Copyright © 2024, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Towards HDR and HFR Video from Rolling-Mixed-Bit Spikings

Towards HDR and HFR Video from Rolling-Mixed-Bit Spikings

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Yakun Chang Yeliduosi Xiaokaiti Yujia Liu Bin Fan Zhaojun Huang Tiejun Huang Boxin Shi Institute of Information Science School of Computer Science Beijing Jiaotong University Beijing Key Laboratory of Advanced Information Science and Network Technology National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University National Engineering Research Center of Visual Technology School of Computer Science Peking University National Key Lab of General AI School of Intelligence Science and Technology Peking University

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

The spiking cameras offer the benefits of high dynamic range (HDR), high temporal resolution, and low data redundancy. However, reconstructing HDR videos in high-speed conditions using single-bit spikings presents challenges due to the limited bit depth. Increasing the bit depth of the spikings is advantageous for boosting HDR performance, but the readout efficiency will be decreased, which is unfavorable for achieving a high frame rate (HFR) video. To address these challenges, we propose a readout mechanism to obtain rolling-mixed-bit (RMB) spikings, which involves inter-leaving multi-bit spikings within the single-bit spikings in a rolling manner, thereby combining the characteristics of high bit depth and efficient readout. Furthermore, we introduce RMB-Net for reconstructing HDR and HFR videos. RMB-Net comprises a cross-bit attention block for fusing mixed-bit spikings and a cross-time attention block for achieving temporal fusion. Extensive experiments conducted on synthetic and real-synthetic data demonstrate the superiority of our method. For instance, pure 3 -bit spikings result in 3 times of data volume, whereas our method achieves comparable performance with less than 2% increase in data volume.

关键词： computer vision Computational modeling Redundancy Cameras Boosting Data models Pattern recognition

来源：评论

学校读者我要写书评

暂无评论

MIXED-ATTENTION AUTO ENCODER FOR MULTI-CLASS INDUSTRIAL ANOMALY DETECTION

arXiv

引用

arXiv 2023年

作者： Liu, Jiangqi Wang, Feng Shanghai Key Laboratory of Multidimensional Information Processing School of Computer Science and Technology East China Normal University Shanghai China

Most existing methods for unsupervised industrial anomaly detection train a separate model for each object category. This kind of approach can easily capture the category-specific feature distributions, but results in high storage cost and low training efficiency. In this paper, we propose a unified mixed-attention auto encoder (MAAE) to implement multi-class anomaly detection with a single model. To alleviate the performance degradation due to the diverse distribution patterns of different categories, we employ spatial attentions and channel attentions to effectively capture the global category information and model the feature distributions of multiple classes. Furthermore, to simulate the realistic noises on features and preserve the surface semantics of objects from different categories which are essential for detecting the subtle anomalies, we propose an adaptive noise generator and a multi-scale fusion module for the pre-trained features. MAAE delivers remarkable performances on the benchmark dataset compared with the state-of-the-art *** Codes 68T45 Copyright © 2023, The Authors. All rights reserved.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

ParCNetV2: Oversized Kernel with Enhanced Attention*

ParCNetV2: Oversized Kernel with Enhanced Attention*

引用

International Conference on computer Vision (ICCV)

作者： Ruihan Xu Haokui Zhang Wenze Hu Shiliang Zhang Xiaoyu Wang National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University Intellifusion Harbin Institute of Technology Shenzhen

Transformers have shown great potential in various computer vision tasks. By borrowing design concepts from transformers, many studies revolutionized CNNs and showed remarkable results. This paper falls in this line of studies. Specifically, we propose a new convolutional neural network, ParCNetV2, that extends the research line of ParCNetV1 by bridging the gap between CNN and ViT. It introduces two key designs: 1) Oversized Convolution (OC) with twice the size of the input, and 2) Bifurcate Gate Unit (BGU) to ensure that the model is input adaptive. Fusing OC and BGU in a unified CNN, ParCNetV2 is capable of flexibly extracting global features like ViT, while maintaining lower latency and better accuracy. Extensive experiments demonstrate the superiority of our method over other convolutional neural networks and hybrid models that combine CNNs and transformers. The code are publicly available at https://***/XuRuihan/ParCNetV2.

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：