CVPR 2021

2,096 papers

Top Research Directions

Research directions most represented by CVPR 2021 papers.

Generalizing from Synthetic Data to the Real World 40
Resolving Depth Ambiguity in Monocular 3D Reconstruction 22
Accelerating Neural Architecture Search Without Full Model Training 20
Mitigating Error Propagation in Pseudo Labeling 19
Global Context Modeling via Non-Local Spatial Attention 18
Disentangling Attribute Control in Generative Models 17
Loss Weighting via Predictive Uncertainty 17
Why Theoretical FLOPs Fail to Predict Real Latency 17
Generative Hallucination for Disoccluded Region Synthesis 17
Deep Learning on Unstructured 3D Point Clouds 17
Preserving Tiny Features Against Spatial Resolution Degradation 16
Reconstructing Complete Geometry from Partial Point Clouds 16
The Sensitivity of Multi-Objective Loss Balancing 16
Cross Task Generalization in Image Restoration 15
Overcoming the Static Scene Assumption 15

Top Authors

AuthorPapers
Luc Van Gool15
Chunhua Shen13
Dacheng Tao12
Chang Xu11
Chen Change Loy11
Ling Shao11
Yi Yang11
Jiaya Jia10
Marc Pollefeys10
Ziwei Liu10
Alan Yuille9
Hongsheng Li9
Jie Zhou9
Yunhe Wang9
Christian Theobalt8
Hongdong Li8
Houqiang Li8
Huchuan Lu8
Lei Zhang8
Leonidas Guibas8

Impactful Papers

Papers from CVPR 2021 that shaped research clusters.

Foundational

Influential

Papers

Title
Omnimatte: Associating Objects and Their Effects in Video
Omni-Supervised Point Cloud Segmentation via Gradual Receptive Field Component Reasoning
One Shot Face Swapping on Megapixels
One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing
One-Shot Neural Ensemble Architecture Search by Diversity-Guided Search Space Shrinking
One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation
On Feature Normalization and Data Augmentation
On Focal Loss for Class-Posterior Probability Estimation: A Theoretical Perspective
On Learning the Geodesic Path for Incremental Learning
Online Learning of a Probabilistic and Adaptive Scene Representation
Online Multiple Object Tracking With Cross-Task Synergy
On Robustness and Transferability of Convolutional Neural Networks
On Self-Contact and Human Pose
On Semantic Similarity in Video Retrieval
On the Difficulty of Membership Inference Attacks
On the Instability of Relative Pose Estimation and RANSAC's Role
On the Integration of Self-Attention and Convolution
OPANAS: One-Shot Path Aggregation Network Architecture Search for Object Detection
Open-Book Video Captioning With Retrieve-Copy-Generate Network
Open-Domain, Content-based, Multi-modal Fact-checking of Out-of-Context Images via Online Resources
Open Domain Generalization with Domain-Augmented Meta-Learning
Opening up Open World Tracking
OpenMix: Reviving Known Knowledge for Discovering Novel Visual Categories in an Open World
OpenRooms: An Open Framework for Photorealistic Indoor Scene Datasets
Open-Set Representation Learning through Combinatorial Embedding
Open-Vocabulary Instance Segmentation via Robust Cross-Modal Pseudo-Labeling
Open-Vocabulary Object Detection Using Captions
Open World Compositional Zero-Shot Learning
Optimal Gradient Checkpoint Search for Arbitrary Computation Graphs
Optimal Quantization Using Scaled Codebook
ORDisCo: Effective and Efficient Usage of Incremental Unlabeled Data for Semi-Supervised Continual Learning
Oriented RepPoints for Aerial Object Detection
Orthogonal Over-Parameterized Training
OSTeC: One-Shot Texture Completion
OTA: Optimal Transport Assignment for Object Detection
OTCE: A Transferability Metric for Cross-Domain Cross-Task Representations
Out-of-Distribution Detection and Generation Using Soft Brownian Offset Sampling and Autoencoders
Out-of-Distribution Detection Using Union of 1-Dimensional Subspaces
Over-the-Air Adversarial Flickering Attacks Against Video Recognition Networks
OW-DETR: Open-world Detection Transformer
PAConv: Position Adaptive Convolution With Dynamic Kernel Assembling on Point Clouds
PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentation
Panoptic-PolarNet: Proposal-Free LiDAR Point Cloud Panoptic Segmentation
Panoramic Image Reflection Removal
Pareidolia Face Reenactment
Pareto Self-Supervised Training for Few-Shot Learning
Parser-Free Virtual Try-On via Distilling Appearance Flows
Part-Aware Panoptic Segmentation
PartGlot: Learning Shape Part Segmentation from Language Reference Games
Partial Feature Selection and Alignment for Multi-Source Domain Adaptation

« prev Page 27/42 next »