CVPR-2022-Papers

官网链接：https://cvpr2022.thecvf.com/

开会时间：2022年6月19日-6月24日

❣❣❣近日，CVPR 2022 接收论文公布！ 总计2067篇！，部分预印版论文也陆续发布中，本文档也将持续收录更新，多多关注!!

❗❗❗ 3月7日更新 14+6 篇.

• 遥感图像融合
  • HyperTransformer: A Textural and Spectral Feature Fusion Transformer for Pansharpening
    ⭐code📰粗解
• 车道线检测
  • Rethinking Efficient Lane Detection via Curve Modeling
    ⭐code📰粗解
• 半监督
  • Class-Aware Contrastive Semi-Supervised Learning
• Deepfake
  • Voice-Face Homogeneity Tells Deepfake
    ⭐code📰粗解
• 其它
  • ACVNet: Attention Concatenation Volume for Accurate and Efficient Stereo Matching
    ⭐code📰粗解
  • Polarity Sampling: Quality and Diversity Control of Pre-Trained Generative Networks via Singular Values
    ⭐code
  • Do Explanations Explain? Model Knows Best
    ⭐code
  • HDNet: High-resolution Dual-domain Learning for Spectral Compressive Imaging
  • E-CIR: Event-Enhanced Continuous Intensity Recovery
    ⭐code
• 目标检测
  • A Versatile Multi-View Framework for LiDAR-based 3D Object Detection with Guidance from Panoptic Segmentation
  • Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving
    ⭐code📰粗解
• 图像合成
  • Interactive Image Synthesis with Panoptic Layout Generation
  • Autoregressive Image Generation using Residual Quantization
    ⭐code📰粗解
• 医学影像
  • Temporal Context Matters: Enhancing Single Image Prediction with Disease Progression Representations

🐦️

• NAS
  • ISNAS-DIP: Image-Specific Neural Architecture Search for Deep Image Prior
• GNN
  • Lifelong Graph Learning
    ⭐code
• 其它
  • Transferability Estimation using Bhattacharyya Class Separability
• 深度估计
  • Toward Practical Self-Supervised Monocular Indoor Depth Estimation
• 人群计数
  • Leveraging Self-Supervision for Cross-Domain Crowd Counting
• 行为预测
  • JRDB-Act: A Large-scale Dataset for Spatio-temporal Action, Social Group and Activity Detection

❗❗❗ 3月4日更新 19 篇.

• HOI
  • HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction
• 跟踪
  • TCTrack: Temporal Contexts for Aerial Tracking
    ⭐code📰粗解
  • Correlation-Aware Deep Tracking
  • Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds
    ⭐code📰粗解
• 目标定位
  • Weakly Supervised Object Localization as Domain Adaption
    ⭐code📰粗解
• 自监督
  • A study on the distribution of social biases in self-supervised learning visual models
• 视频
  • Efficient Video Instance Segmentation via Tracklet Query and Proposal
    🏠project📺video📰粗解
• 三维
  • LGT-Net: Indoor Panoramic Room Layout Estimation with Geometry-Aware Transformer Network
    ⭐code📰粗解
• 其它
  • PINA: Learning a Personalized Implicit Neural Avatar from a Single RGB-D Video Sequence
    ⭐code🏠project📺video📰粗解
  • CAFE: Learning to Condense Dataset by Aligning Features
    ⭐code📰粗解
  • Enhancing Adversarial Robustness for Deep Metric Learning
  • BatchFormer: Learning to Explore Sample Relationships for Robust Representation Learning
    ⭐code📰粗解
• 光场
  • Occlusion-Aware Cost Constructor for Light Field Depth Estimation
    ⭐code📰粗解
• 图像到图像翻译
  • Exploring Patch-wise Semantic Relation for Contrastive Learning in Image-to-Image Translation Tasks
• 分割
  • SoftGroup for 3D Instance Segmentation on Point Clouds
    ⭐code📰粗解
  • Bending Reality: Distortion-aware Transformers for Adapting to Panoramic Semantic Segmentation
    ⭐code📰粗解
• 深度估计
  • NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimation
• 数据增强
  • 3D Common Corruptions and Data Augmentation
    ⭐code🏠project📺video📰粗解
• 对抗
  • Label-Only Model Inversion Attacks via Boundary Repulsion

目录

🐱	🐶	🐯	🐺
1.其它	2.Image Segmentation(图像分割)	3.Image Progress(图像处理)	4.Image Captioning(图像字幕)
5.Object Detection(目标检测)	6.Object Tracking(目标跟踪)	7.Point Cloud(点云)	8.Action Detection(人体动作检测与识别)
9.Human Pose Estimation(人体姿态估计)	10.3D(三维视觉)	11.Face	12.Image-to-Image Translation(图像到图像翻译)
13.GAN	14.Video	15.Transformer	16.Semi/self-supervised learning(半/自监督)
17.Medical Image(医学影像)	18.Person Re-Identification(人员重识别)

Light Field(光场)

• Occlusion-Aware Cost Constructor for Light Field Depth Estimation
  ⭐code📰粗解

Adversarial Learning(对抗学习)

• 对抗样本
  • Label-Only Model Inversion Attacks via Boundary Repulsion

Human-Object Interaction(人物交互)

• HOI4D: A 4D Egocentric Dataset for Category-Level Human-Object Interaction

数据增强

• 🐦️AlignMix: Improving representation by interpolating aligned features
• 3D Common Corruptions and Data Augmentation
  ⭐code🏠project📺video📰粗解

GCN/GNN

• GNN
  • 🐦️Lifelong Graph Learning
    ⭐code

Image Synthesis/Generation(图像合成)

• Interactive Image Synthesis with Panoptic Layout Generation
• Autoregressive Image Generation using Residual Quantization
  ⭐code📰粗解

UAV/Remote Sensing/Satellite Image(无人机/遥感/卫星图像)

• 遥感图像融合
  • HyperTransformer: A Textural and Spectral Feature Fusion Transformer for Pansharpening
    ⭐code📰粗解

Autonomous vehicles(自动驾驶)

• 车道线检测
  • Rethinking Efficient Lane Detection via Curve Modeling
    ⭐code📰粗解
• 行为预测
  • 🐦️JRDB-Act: A Large-scale Dataset for Spatio-temporal Action, Social Group and Activity Detection

Neural Architecture Search(神经架构搜索)

• 🐦️ISNAS-DIP: Image-Specific Neural Architecture Search for Deep Image Prior

18.Person Re-Identification(人员重识别)

• 人群计数
  • Leveraging Self-Supervision for Cross-Domain Crowd Counting

17.Medical Image(医学影像)

• Temporal Context Matters: Enhancing Single Image Prediction with Disease Progression Representations

16.Semi/self-supervised learning(半/自监督)

• 自监督
  • A study on the distribution of social biases in self-supervised learning visual models
• 半监督
  • Class-Aware Contrastive Semi-Supervised Learning

15.Transformer

• Fast Point Transformer

14.Video

• 动作分割
  • Unsupervised Activity Segmentation by Joint Representation Learning and Online Clustering
    📺video
• 视频实例分割(VIS)
  • Efficient Video Instance Segmentation via Tracklet Query and Proposal
    🏠project📺video📰粗解

13.GAN

• 🐦️HyperInverter: Improving StyleGAN Inversion via Hypernetwork
  🏠project

12.Image-to-Image Translation(图像到图像翻译)

• Exploring Patch-wise Semantic Relation for Contrastive Learning in Image-to-Image Translation Tasks

11.Face

• 人脸
  • Protecting Celebrities with Identity Consistency Transformer
• Deepfake
  • Voice-Face Homogeneity Tells Deepfake
    ⭐code📰粗解

10.3D(三维视觉)

• 深度估计
  • OmniFusion: 360 Monocular Depth Estimation via Geometry-Aware Fusion
  • NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimation
  • 🐦️Toward Practical Self-Supervised Monocular Indoor Depth Estimation
• 房间布局
  • LGT-Net: Indoor Panoramic Room Layout Estimation with Geometry-Aware Transformer Network
    ⭐code📰粗解

9.Human Pose Estimation(人体姿态估计)

• 3D pose
  • MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video
• 4D 人体捕获
  • H4D: Human 4D Modeling by Learning Neural Compositional Representation

8.Action Detection(人体动作检测与识别)

• 动作检测
  • Colar: Effective and Efficient Online Action Detection by Consulting Exemplars

7.Point Cloud(点云)

• 3D 点云
  • CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding
    ⭐code📰粗解
    CrossPoint，一个用于 3D 点云表征学习的简单自监督学习框架。虽然该方法是在合成的三维物体数据集上训练的，但在下游任务中的实验结果，如三维物体分类和三维物体部分分割，在合成和真实世界的数据集中都证明了该方法在学习可迁移表征方面的有效性。
  • A Unified Query-based Paradigm for Point Cloud Understanding

6.Object Tracking(目标跟踪)

• TCTrack: Temporal Contexts for Aerial Tracking
  ⭐code📰粗解
• Correlation-Aware Deep Tracking
• 3D 目标跟踪
  • Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds
    ⭐code📰粗解

5.Object Detection(目标检测)

• DN-DETR: Accelerate DETR Training by Introducing Query DeNoising
  ⭐code📰粗解
• 目标定位
  • Weakly Supervised Object Localization as Domain Adaption
    ⭐code📰粗解
• 3D
  • A Versatile Multi-View Framework for LiDAR-based 3D Object Detection with Guidance from Panoptic Segmentation
  • Pseudo-Stereo for Monocular 3D Object Detection in Autonomous Driving
    ⭐code📰粗解

4.Image Captioning(图像字幕)

• 字幕
  • X-Trans2Cap: Cross-Modal Knowledge Transfer using Transformer for 3D Dense Captioning

3.Image Progress(图像处理)

• 图像修复
  • Incremental Transformer Structure Enhanced Image Inpainting with Masking Positional Encoding
    ⭐code📰粗解

2.Image Segmentation(图像分割)

• 实例分割
  • 3D 实例分割
    • SoftGroup for 3D Instance Segmentation on Point Clouds
      ⭐code📰粗解
  • 🐦️FreeSOLO: Learning to Segment Objects without Annotations
• 语义分割
  • Bending Reality: Distortion-aware Transformers for Adapting to Panoramic Semantic Segmentation
    ⭐code📰粗解
  • 弱监督语义分割
    • Class Re-Activation Maps for Weakly-Supervised Semantic Segmentation
      ⭐code📰粗解

1.其它

• Instance-wise Occlusion and Depth Orders in Natural Scenes
• IFOR: Iterative Flow Minimization for Robotic Object Rearrangement
  🏠project
• PINA: Learning a Personalized Implicit Neural Avatar from a Single RGB-D Video Sequence
  ⭐code🏠project📺video📰粗解
• CAFE: Learning to Condense Dataset by Aligning Features
  ⭐code📰粗解
• Enhancing Adversarial Robustness for Deep Metric Learning
• BatchFormer: Learning to Explore Sample Relationships for Robust Representation Learning
  ⭐code📰粗解
• ACVNet: Attention Concatenation Volume for Accurate and Efficient Stereo Matching
  ⭐code📰粗解
• Polarity Sampling: Quality and Diversity Control of Pre-Trained Generative Networks via Singular Values
  ⭐code
• Do Explanations Explain? Model Knows Best
  ⭐code
• HDNet: High-resolution Dual-domain Learning for Spectral Compressive Imaging
• E-CIR: Event-Enhanced Continuous Intensity Recovery
  ⭐code
• 🐦️Transferability Estimation using Bhattacharyya Class Separability

论文尚未公布

AxIoU: An Axiomatically Justified Measure for Video Moment Retrieval

ID:Cyelie multi-Variate Function for self-supervised image denoising by disentangling noise form image

Habitat-Web: Learning Embodied Object-Search Strategies from Human Demonstrations at Scale

Diverse Plausible 360-Degree Image Outpainting for Efficient 3DCG Background Creation

来源
[Two Systems in Thinking: Dual-System Transformer for Grounded Situation Recognition]
[Autoregressive Image Generation using Residual Quantization]
✔️Instance-wise Occlusion and Depth Orders in Natural Scenes
[Style Neophile: Constantly Seeking Novel Styles for Domain Generalization]
[ReSTR: Convolution-free Referring Image Segmentation Using Transformers]
[FIFO: Learning Fog-invariant Features for Foggy Scene Segmentation]
[TransforMatcher: Match-to-Match Attention for Semantic Correspondence]
[Reflection and Rotation Symmetry Detection via Equivariant Learning]
[Semi-supervised Semantic Segmentation with Error Localization Network]
[Future Transformer for Long-term Action Anticipation]
[Self-Taught Metric Learning without Labels]
✔️Fast Point Transformer
[Integrative Few-Shot Learning for Classification and Segmentation]
[Scene Painting via Semantic Image Synthesis]
[Detector-Free Weakly Supervised Group Activity Recognition]

扫码CV君微信（注明：CVPR）入微信交流群：