2022 IEEE International Conference on Visual Communications and Image Processing (VCIP)最新文献_第3页

Mining Regional Relation from Pixel-wise Annotation for Scene Parsing 面向场景分析的逐像素标注区域关系挖掘

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008859

Zichen Song, Hongliang Li, Heqian Qiu, Xiaoliang Zhang

{"title":"Mining Regional Relation from Pixel-wise Annotation for Scene Parsing","authors":"Zichen Song, Hongliang Li, Heqian Qiu, Xiaoliang Zhang","doi":"10.1109/VCIP56404.2022.10008859","DOIUrl":"https://doi.org/10.1109/VCIP56404.2022.10008859","url":null,"abstract":"Scene parsing is an important and challenging task in computer vision, which assigns semantic labels to each pixel in the entire scene. Existing scene parsing methods only utilize pixel-wise annotation as the supervision of neural network, thus, some similar categories are easy to be misclassified in the complex scenes without the utilization of regional relation. To tackle these above challenging problems, a Regional Relation Network (RRNet) is proposed in this paper, which aims to boost the scene parsing performance by mining regional relation from pixel-wise annotation. Specifically, the pixel-wise annotation is divided into a lot of fixed regions, so that intra- and inter-regional relation are able to be extracted as the supervision of network. We firstly design an intra-regional relation module to predict category distribution in each fixed region, which is helpful for reducing the misclassification phenomenon in regions. Secondly, an inter-regional relation module is proposed to learn the relationships among each region in scene images. With the guideline of relation information extracted from the ground truth, the network is able to learn more discriminative relation representations. To validate our proposed model, we conduct experiments on three typical datasets, including NYU-depth-v2, PASCAL-Context and ADE20k. The achieved competitive results on all three datasets demonstrate the effectiveness of our method.","PeriodicalId":269379,"journal":{"name":"2022 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"728 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126948980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Identify, Guess and Reconstruct: Three Principles for Cloud Removal Task 识别、猜测和重构:云移除任务的三个原则

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008852

Sibo Wu, Mengqiu Xu, Ming Wu, Chuang Zhang, Hua Shen

{"title":"Identify, Guess and Reconstruct: Three Principles for Cloud Removal Task","authors":"Sibo Wu, Mengqiu Xu, Ming Wu, Chuang Zhang, Hua Shen","doi":"10.1109/VCIP56404.2022.10008852","DOIUrl":"https://doi.org/10.1109/VCIP56404.2022.10008852","url":null,"abstract":"Remote sensing images serve a significant role in earth observation to tackle climate change and post-disaster reconstruction concerns. However, optical images are obscured by clouds or haze, preventing precise earth observation; hence, cloud removal has been a hot topic among concerned scholars. The objective of this article is to make cloud removal more efficient and explicable by proposing three principles: identifying clouds, guessing objects beneath the clouds, and reconstructing the cloudy area. In addition, a modified dual contrastive learning Generative Adversarial Network is proposed based on these three principles by adding cloud detection and weight sharing strategy to obtain cloud semantics. In particular, we align two datasets by forming a quaternary sample pair that includes not only optical pictures and SAR images, but also region information for a more precise reconstruction. Our experiment results on the integrated dataset reveal the superiority of proposed method over previous cloud removal methods and the effectiveness of added modules through ablation experiments, with PSNR and SSIM values of 26.2 and 0.728, respectively.","PeriodicalId":269379,"journal":{"name":"2022 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122562751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Real-time Learned Image Codec on FPGA 基于FPGA的实时学习图像编解码器

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008876

Heming Sun, Qingyang Yi, Fangzheng Lin, Lu Yu, J. Katto, M. Fujita

引用次数: 0

Tire Pattern Image Classification using Variational Auto-Encoder with Contrastive Learning 基于对比学习的变分自编码器轮胎图案图像分类

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008835

Jianning Yang, Jiahao Xue, Xiao Feng, Chaoqi Song, Yu Hao

引用次数: 0

A Lightweight Tire Tread Image Classification Network 一种轻型轮胎胎面图像分类网络

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008894

Fenglei Zhang, Da Li, Shenghua Li, Weili Guan, Meng Liu

引用次数: 0

SparseARFM-SI: Rotary Point Cloud Place Recognition Based on Multi-Resolution and Attention Mechanism SparseARFM-SI:基于多分辨率和注意机制的旋转点云位置识别

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008851

Xiaohang Liu, Ruonan Zhang, Jie Wang, Ge Li

引用次数: 0

Block Importance Mapping for Video Encoding 视频编码的块重要性映射

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008857

Jack Enhorn, Christopher Hollmann, Rickard Sjöberg, Jacob Ström, Per Wennersten

引用次数: 0

Rate Controllable Learned Image Compression Based on RFL Model 基于RFL模型的速率可控学习图像压缩

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008802

Saiping Zhang, Luge Wang, Xionghui Mao, Fuzheng Yang, Shuai Wan

引用次数: 0

Refine-PU: A Graph Convolutional Point Cloud Upsampling Network using Spatial Refinement 一个使用空间细化的图卷积点云上采样网络

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008811

Yilin Liu, Yumei Wang, Yu Liu

{"title":"Refine-PU: A Graph Convolutional Point Cloud Upsampling Network using Spatial Refinement","authors":"Yilin Liu, Yumei Wang, Yu Liu","doi":"10.1109/VCIP56404.2022.10008811","DOIUrl":"https://doi.org/10.1109/VCIP56404.2022.10008811","url":null,"abstract":"Upsampling of 3D point clouds plays an important role in point cloud reconstruction, rendering, meshing and analysis. Most of the existing point cloud upsampling networks are three-tier cascaded networks that use a combination of feature extraction, feature expansion, and coordinate reconstruction. However, for the point clouds reconstructed by this network architecture, there is a non-negligible deviation between the point clouds after upsampling and the ground truth, especially in the details. In this paper, we propose a four-tier cascaded graph convolutional network architecture called Refine-PU, which di-vides the network into four parts, i.e., feature extraction, feature expansion, coordinate reconstruction, and spatial refinement. We first design a multi-scale graph convolutional feature extractor called Dense Edge Conv (DEC) to better capture the global and the local structural features of point clouds. In addition, the network appends a spatial refinement module on the basis of the traditional three-tier cascaded network, in order to further adjust the details of the coarse dense point clouds obtained after upsampling to make them more consistent with the ground truth. Extensive experiments implemented on both synthetic and real-scanned datasets demonstrate the superiority of our method over the state-of-the-art methods both quantitatively and qualitatively.","PeriodicalId":269379,"journal":{"name":"2022 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127418121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Global Homography Motion Compensation for Versatile Video Coding 通用视频编码的全局单应性运动补偿

2022 IEEE International Conference on Visual Communications and Image Processing (VCIP) Pub Date : 2022-12-13 DOI: 10.1109/VCIP56404.2022.10008901

Yao Li, Zhuoyuan Li, Li Li, Dong Liu, Houqiang Li

{"title":"Global Homography Motion Compensation for Versatile Video Coding","authors":"Yao Li, Zhuoyuan Li, Li Li, Dong Liu, Houqiang Li","doi":"10.1109/VCIP56404.2022.10008901","DOIUrl":"https://doi.org/10.1109/VCIP56404.2022.10008901","url":null,"abstract":"In Versatile Video Coding (VVC), local affine motion compensation (LAMC) is adopted to handle complex motions, such as rotation and zooming. However, it is inefficient to use LAMC to handle the global motion due to the following two reasons. First, the use of LAMC may lead to some extra bit cost on the affine motion model parameters. Second, the precision of LAMC is restricted by the MV precision of the control points. Therefore, in this paper, we propose a global homography motion compensation (GHMC) framework to better characterize the global motion. For each coding block, an extra mode is added to perform motion compensation based on an 8-parameter global homography motion model. In addition, an extrapolation scheme is designed to derive the parameters from reference frames to save the bit cost for signaling them. The proposed framework is implemented into the VVC reference software VTM-6.0. Experimental results show that, on average, 0.69% and 0.66% BD-rate reduction is achieved under Low Delay P and Low Delay B configurations, respectively, for sequences with rich complex global motions.","PeriodicalId":269379,"journal":{"name":"2022 IEEE International Conference on Visual Communications and Image Processing (VCIP)","volume":"168 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126965221","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1