CVPR 2022

2,477 papers

Top Research Directions

Research directions most represented by CVPR 2022 papers.

Generalizing from Synthetic Data to the Real World 41
Overcoming Local Receptive Field Constraints in Visual Models 38
Scaling Visual Attention Beyond Quadratic Complexity 36
Mitigating Error Propagation in Pseudo Labeling 28
Disentangling Attribute Control in Generative Models 23
Bypassing the Computational Bottleneck of Adversarial Training 23
Adversarial Training for Domain Invariant Representations 20
Addressing Inductive Bias Deficiency in Vision Transformers 19
Probabilistic Trajectory Forecasting for Autonomous Systems 18
Resolving Geometric Symmetry Ambiguity in 3D Pose Estimation 18
Decoupling Geometry and Appearance in Generative Models 17
Preserving Tiny Features Against Spatial Resolution Degradation 17
Resolving Depth Ambiguity in Monocular 3D Reconstruction 16
Cross Task Generalization in Image Restoration 15
The Representational Bottleneck of Frozen Foundation Models 15

Top Authors

AuthorPapers
Luc Van Gool24
Dacheng Tao20
Alan Yuille12
Dongdong Chen12
Jiaya Jia12
Antonio Torralba11
Jie Zhou11
Bernt Schiele10
Cewu Lu10
Errui Ding10
Jitendra Malik10
Yanwei Fu10
Yi Yang10
Chen Change Loy9
Fahad Shahbaz Khan9
Leonidas Guibas9
Minsu Cho9
Qi Tian9
Xian-Sheng Hua9
Xiaodan Liang9

Impactful Papers

Papers from CVPR 2022 that shaped research clusters.

Foundational

Influential

Papers

Title
The Neurally-Guided Shape Parser: Grammar-based Labeling of 3D Shape Regions with Approximate Inference
The Norm Must Go On: Dynamic Unsupervised Domain Adaptation by Normalization
'The Pedestrian next to the Lamppost" Adaptive Object Graphs for Better Instantaneous Mapping
The Principle of Diversity: Training Stronger Vision Transformers Calls for Reducing All Levels of Redundancy
The Probabilistic Normal Epipolar Constraint for Frame- To-Frame Rotation Optimization under Uncertain Feature Positions
The Two Dimensions of Worst-case Training and Their Integrated Effect for Out-of-domain Generalization
The Wanderings of Odysseus in 3D Scenes
Think Global, Act Local: Dual-scale Graph Transformer for Vision-and-Language Navigation
Think Twice Before Detecting GAN-generated Fake Images from their Spectral Domain Imprints
Thin-Plate Spline Motion Model for Image Animation
Time3D: End-to-End Joint Monocular 3D Object Detection and Tracking for Autonomous Driving
Time Lens++: Event-based Frame Interpolation with Parametric Nonlinear Flow and Multi-scale Fusion
TimeReplayer: Unlocking the Potential of Event Cameras for Video Interpolation
TINC: Tree-Structured Implicit Neural Compression
TO-FLOW: Efficient Continuous Normalizing Flows with Temporal Optimization adjoint with Moving Speed
TopFormer: Token Pyramid Transformer for Mobile Semantic Segmentation
Topologically-Aware Deformation Fields for Single-View 3D Reconstruction
Topology Preserving Local Road Network Estimation from Single Onboard Camera Image
Topology-Preserving Shape Reconstruction and Registration via Neural Diffeomorphic Flow
Total Variation Optimization Layers for Computer Vision
Toward Fast, Flexible, and Robust Low-Light Image Enhancement
Toward Practical Monocular Indoor Depth Estimation
Towards All-in-One Pre-Training via Maximizing Multi-Modal Mutual Information
Towards An End-to-End Framework for Flow-Guided Video Inpainting
Towards Better Plasticity-Stability Trade-off in Incremental Learning: A Simple Linear Connector
Towards Better Understanding Attribution Methods
Towards Bidirectional Arbitrary Image Rescaling: Joint Optimization and Cycle Idempotence
Towards Bridging the Performance Gaps of Joint Energy-based Models
Towards Compositional Adversarial Robustness: Generalizing Adversarial Training to Composite Semantic Perturbations
Towards Data-Free Model Stealing in a Hard Label Setting
Towards Discovering the Effectiveness of Moderately Confident Samples for Semi-Supervised Learning
Towards Diverse and Natural Scene-aware 3D Human Motion Synthesis
Towards Driving-Oriented Metric for Lane Detection Models
Towards Efficient and Scalable Sharpness-Aware Minimization
Towards Efficient Data Free Blackbox Adversarial Attack
Towards efficient feature sharing in MIMO architectures
Towards Efficient Use of Multi-Scale Features in Transformer-Based Object Detectors
Towards End-to-End Unified Scene Text Detection and Layout Analysis
Towards Fast Adaptation of Pretrained Contrastive Models for Multi-channel Video-Language Retrieval
Towards Fewer Annotations: Active Learning via Region Impurity and Prediction Uncertainty for Domain Adaptive Semantic Segmentation
Towards General Purpose Vision Systems: An End-to-End Task-Agnostic Vision-Language Architecture
Towards Implicit Text-Guided 3D Shape Generation
Towards Language-Free Training for Text-to-Image Generation
Towards Layer-wise Image Vectorization
Towards Low-Cost and Efficient Malaria Detection
Towards Multi-domain Single Image Dehazing via Test-time Training
Towards Multimodal Depth Estimation from Light Fields
Towards Noiseless Object Contours for Weakly Supervised Semantic Segmentation
Towards Open-Set Object Detection and Discovery
Towards Practical Certifiable Patch Defense with Vision Transformer

« prev Page 45/50 next »