计算机视觉论文速递

编辑 | | AI速递

计算机视觉 3月19日

[1] Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild

标题：野外交互式视频对象分割中推荐帧的学习

链接：https://arxiv.org/abs/2103.10391

[2] Investigate Indistinguishable Points in Semantic Segmentation of 3D Point Cloud

标题：三维点云语义分割中不可区分点的研究

链接：https://arxiv.org/abs/2103.10339

[3] SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation

标题：SG-NET：一级视频实例分割的空间粒度网络

链接：https://arxiv.org/abs/2103.10284

[4] A Location-Sensitive Local Prototype Network for Few-Shot Medical Image Segmentation

标题：一种位置敏感的局部原型网络用于Few-Shot医学图像分割

链接：https://arxiv.org/abs/2103.10178

[5] On Semantic Similarity in Video Retrieval

标题：视频检索中的语义相似度研究

链接：https://arxiv.org/abs/2103.10095

[6] Discriminative and Semantic Feature Selection for Place Recognition towards Dynamic Environments

标题：面向动态环境的地点识别的区别性和语义特征选择

链接：https://arxiv.org/abs/2103.10023

[7] Topology-Aware Segmentation Using Discrete Morse Theory

标题：基于离散Morse理论的拓扑感知分割

链接：https://arxiv.org/abs/2103.09992

[8] Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networks
标题：神经零件：使用可逆神经网络学习富有表现力的3D形状抽象
链接 | https://arxiv.org/abs/2103.10429

[9] Large Scale Image Completion via Co-Modulated Generative Adversarial Networks
标题：通过协同调制的对抗网络进行大规模图像完成
链接 | https://arxiv.org/abs/2103.10428

[10] Using latent space regression to analyze and leverage compositionality in GANs
标题：使用潜在空间回归分析和利用GAN中的组成
链接 | https://arxiv.org/abs/2103.10426

[11] Computer Vision Aided URLL Communications: Proactive Service Identification and Coexistence
标题：基于离散Morse理论的拓扑感知分割
链接 | https://arxiv.org/abs/2103.10419
项目链接 | https://viwi-dataset.net/

[12] Learning to Recommend Frame for Interactive Video Object Segmentation in the Wild
标题：学习在野外推荐用于交互式视频对象分割的帧
链接 | https://arxiv.org/abs/2103.10391

[13] Challenges of 3D Surface Reconstruction in Capsule Endoscopy
标题：胶囊内窥镜中3D表面重建的挑战
链接 | https://arxiv.org/abs/2103.10390

[14] FastNeRF: High-Fidelity Neural Rendering at 200FPS
标题：FastNeRF：200FPS的高保真神经渲染
链接 | https://arxiv.org/abs/2103.10380

[15] Consistency-based Active Learning for Object Detection
标题：基于一致性的主动学习的对象检测
链接 | https://arxiv.org/abs/2103.10374

[16] Investigate Indistinguishable Points in Semantic Segmentation of 3D Point Cloud
标题：在3D点云的语义分割中研究难以区分的点
链接 | https://arxiv.org/abs/2103.10339

[17] SG-Net: Spatial Granularity Network for One-Stage Video Instance Segmentation
标题：SG-Net：用于一阶段视频实例分割的空间粒度网络
链接 | https://arxiv.org/abs/2103.10284

[18] Collective Decision of One-vs-Rest Networks for Open Set Recognition
标题：开放集识别的一对多静止网络的集体决策
链接 | https://arxiv.org/abs/2103.10230

[19] Learning to Amend Facial Expression Representation via De-albino and Affinity
标题：学习通过De-albino和Affinity修改面部表情表示
链接 | https://arxiv.org/abs/2103.10189

[20] OmniPose: A Multi-Scale Framework for Multi-Person Pose Estimation
标题：OmniPose：用于多人姿势估计的多尺度框架
链接 | https://arxiv.org/abs/2103.10180

[21] Spectral Reconstruction and Disparity from Spatio-Spectrally Coded Light Fields via Multi-Task Deep Learning
标题：通过多任务深度学习对空间谱编码光场进行谱重构和视差
链接 | https://arxiv.org/abs/2103.10179

[22] A Location-Sensitive Local Prototype Network for Few-Shot Medical Image Segmentation
标题：用于少量医学图像分割的位置敏感型本地原型网络
链接 | https://arxiv.org/abs/2103.10178

[23] Sequential End-to-end Network for Efficient Person Search
标题：顺序的端到端网络，以进行有效的人员搜索
链接 | https://arxiv.org/abs/2103.10148

[24] KoDF: A Large-scale Korean DeepFake Detection Dataset
标题：KoDF：大规模韩国DeepFake检测数据集
链接 | https://arxiv.org/abs/2103.10094

[25] Which to Match? Selecting Consistent GT-Proposal Assignment for Pedestrian Detection
标题：匹配哪一个？为行人检测选择一致的GT提案分配
链接 | https://arxiv.org/abs/2103.10091
备注 | This manuscript is waiting for further improvement

[26] Enhancing Transformer for Video Understanding Using Gated Multi-Level Attention and Temporal Adversarial Training
标题：使用门控多级注意力和时间对抗训练增强用于视频理解的变压器
链接 | https://arxiv.org/abs/2103.10043

[27] SparsePoint: Fully End-to-End Sparse 3D Object Detector
标题：[38] SparsePoint：全面的端到端稀疏3D对象检测器
链接 | https://arxiv.org/abs/2103.10042

[28] RangeDet:In Defense of Range View for LiDAR-based 3D Object Detection
标题：RangeDet：用于基于LiDAR的3D对象检测的范围视图防御
链接 | https://arxiv.org/abs/2103.10039

[29] Impressions2Font: Generating Fonts by Specifying Impressions
标题：Impressions2Font：通过指定印象生成字体
链接 | https://arxiv.org/abs/2103.10036

[30] Discriminative and Semantic Feature Selection for Place Recognition towards Dynamic Environments
标题：动态环境中位置识别的语义区分和语义选择
链接 | https://arxiv.org/abs/2103.10023

[31] Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE
标题：使用分层VQ-VAE生成图像修复的多样结构
链接 | https://arxiv.org/abs/2103.10022

[32] Rapid treatment planning for low-dose-rate prostate brachytherapy with TP-GAN
标题：使用TP-GAN进行低剂量率前列腺近距离放射治疗的快速治疗计划
链接 | https://arxiv.org/abs/2103.09996

[33] Topology-Aware Segmentation Using Discrete Morse Theory
标题：离散莫尔斯理论的拓扑感知分割
链接 | https://arxiv.org/abs/2103.09992

[34] Deep Wiener Deconvolution: Wiener Meets Deep Learning for Image Deblurring
标题：深度维纳反卷积：维纳遇到深度学习以进行图像去模糊
链接 | https://arxiv.org/abs/2103.09962

[35] Learning to Resize Images for Computer Vision Tasks
标题：学习调整图像大小以执行计算机视觉任务
链接 | https://arxiv.org/abs/2103.09950

[36] Fast and High-Quality Blind Multi-Spectral Image Pansharpening
标题：快速，高质量的盲多光谱图像锐化
链接 | https://arxiv.org/abs/2103.09943

[37] Impact of Facial Tattoos and Paintings on Face Recognition Systems
标题：面部纹身和绘画对面部识别系统的影响
链接 | https://arxiv.org/abs/2103.09939

[38] MSMatch: Semi-Supervised Multispectral Scene Classification with Few Labels
标题：具有少量标签的半监督多光谱场景分类
链接 : https://arxiv.org/abs/2103.10368

[39] DanceNet3D: Music Based Dance Generation with Parametric Motion Transformer
标题：DanceNet3D：带参数运动变压器的基于音乐的舞蹈生成
链接 : https://arxiv.org/abs/2103.10206

[40] Bias-Free FedGAN
标题：基于离散Morse理论的拓扑感知分割
链接 : https://arxiv.org/abs/2103.09876