2021年3月26最新论文（计算机视觉，图像分割，图像识别，图像分类）

[1]-----------------------------------------------------[25]

[26] Generative-Adversarial-Networks-based Ghost Recognition
基于生成-专家网络的幽灵识别
链接 | https://arxiv.org/abs/2103.13858

[27] Hierarchical Deep CNN Feature Set-Based Representation Learning for Robust Cross-Resolution Face Recognition
基于分层深度CNN特征集的表示学习，用于鲁棒的交叉分辨率人脸识别
链接 | https://arxiv.org/abs/2103.13851

[28] Universal Representation Learning from Multiple Domains for Few-shot Classification
从多个领域进行通用表示学习以进行少量分类
链接 | https://arxiv.org/abs/2103.13841

[29] Multi-frame Super-resolution from Noisy Data
噪声数据的多帧超分辨率
链接 | https://arxiv.org/abs/2103.13778

[30] Patch Craft: Video Denoising by Deep Modeling and Patch Matching
补丁程序：通过深度建模和补丁匹配进行视频降噪
链接 | https://arxiv.org/abs/2103.13767

[31] I^3Net: Implicit Instance-Invariant Network for Adapting One-Stage Object Detectors
I ^ 3Net：用于适应一阶段对象检测器的隐式实例不变网络
链接 | https://arxiv.org/abs/2103.13757

[32] Video Instance Segmentation with a Propose-Reduce Paradigm
使用Propose-Reduce范式进行视频实例分割
链接 | https://arxiv.org/abs/2103.13746

[33] KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPs
KiloNeRF：使用成千上万的微型MLP加速神经辐射场
链接 | https://arxiv.org/abs/2103.13744

[34] Spirit Distillation: Precise Real-time Prediction with Insufficient Data
酒精蒸馏：数据不足时的精确实时预测
链接 | https://arxiv.org/abs/2103.13733

[35] GyroFlow: Gyroscope-Guided Unsupervised Optical Flow Learning
GyroFlow：陀螺仪引导的无监督光流学习
链接 | https://arxiv.org/abs/2103.13725

[36] AttrLostGAN: Attribute Controlled Image Synthesis from Reconfigurable Layout and Style
AttrLostGAN：通过可重新配置的布局和样式进行属性控制的图像合成
链接 | https://arxiv.org/abs/2103.13722

[37] Vectorization and Rasterization: Self-Supervised Learning for Sketch and Handwriting
矢量化和光栅化：素描和手写的自我指导学习
链接 | https://arxiv.org/abs/2103.13716

[39] SSLayout360: Semi-Supervised Indoor Layout Estimation from 360∘ Panorama
SSLayout360：360°全景图的半监督室内布局估计
链接 | https://arxiv.org/abs/2103.13696

[40] MBA-VO: Motion Blur Aware Visual Odometry
MBA-VO：运动模糊感知视觉里程表
链接 | https://arxiv.org/abs/2103.13684

[41] JDSR-GAN: Constructing A Joint and Collaborative Learning Network for Masked Face Super-Resolution
JDSR-GAN：构建联合协作学习网络以实现蒙面超分辨率
链接 | https://arxiv.org/abs/2103.13676

[42] Contrast to Divide: Self-Supervised Pre-Training for Learning with Noisy Labels
差异对比：带有噪声标签的自学式预培训
链接 | https://arxiv.org/abs/2103.13646

[43] Asymmetric CNN for image super-resolution
非对称CNN用于图像超分辨率
链接 | https://arxiv.org/abs/2103.13634

[44] A Survey of Quantization Methods for Efficient Neural Network Inference
高效神经网络推理的量化方法研究
链接 | https://arxiv.org/abs/2103.13630

[45] Learning Probabilistic Ordinal Embeddings for Uncertainty-Aware Regression
学习概率序数嵌入以进行不确定性感知回归
链接 | https://arxiv.org/abs/2103.13629

[46] Gaussian Guided IoU: A Better Metric for Balanced Learning on Object Detection
高斯制导IoU：一种更好的衡量目标检测平衡的指标
链接 | https://arxiv.org/abs/2103.13613

[47] THAT: Two Head Adversarial Training for Improving Robustness at Scale
那：两次头对抗训练，以提高规模的鲁棒性
链接 | https://arxiv.org/abs/2103.13612

[48] Exploiting Class Similarity for Machine Learning with Confidence Labels and Projective Loss Functions
利用置信标签和投影损失函数开发机器学习的类相似性
链接 | https://arxiv.org/abs/2103.13607

[49] Learning Dynamic Alignment via Meta-filter for Few-shot Learning
通过元过滤器学习动态对齐，以进行少量学习
链接 | https://arxiv.org/abs/2103.13582

[50] STA-VPR: Spatio-temporal Alignment for Visual Place Recognition
STA-VPR：用于视觉位置识别的时空对齐
链接 | https://arxiv.org/abs/2103.13580

[51] Test-Time Training for Deformable Multi-Scale Image Registration
变形多尺度图像配准的即时培训
链接 | https://arxiv.org/abs/2103.13578

[52] MetaAlign: Coordinating Domain Alignment and Classification for Unsupervised Domain Adaptation
MetaAlign：无监督域自适应的协调域对齐和分类
链接 | https://arxiv.org/abs/2103.13575

[53] Deepfake Forensics via An Adversarial Game
通过对抗游戏进行的Deepfake取证
链接 | https://arxiv.org/abs/2103.13567

[54] On Evolving Attention Towards Domain Adaptation
关于领域适应的发展注意力
链接 | https://arxiv.org/abs/2103.13561

[55] Rethinking Self-Supervised Learning: Small is Beautiful
重新思考自我监督学习：小就是美
链接 | https://arxiv.org/abs/2103.13559