CVPR 2021 accepted papers on Detection and segmentation
289 of the 1,661 papers accepted at CVPR 2021 (17.4%) are on Detection and segmentation, compared with 16.6% at CVPR 2020. They fall into 5 topics.
See CVPR 2021 in the interactive map
Submitting to CVPR? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
CNN features · 110 papers
- Coordinate Attention for Efficient Mobile Network Design (6,279 citations)
- Rethinking Semantic Segmentation From a Sequence-to-Sequence Perspective With Transformers (3,682 citations)
- LoFTR: Detector-Free Local Feature Matching With Transformers (1,683 citations)
- Scaled-YOLOv4: Scaling Cross Stage Partial Network (1,672 citations)
- Sparse R-CNN: End-to-End Object Detection With Learnable Proposals (1,489 citations)
- Bottleneck Transformers for Visual Recognition (1,304 citations)
- VarifocalNet: An IoU-Aware Dense Object Detector (1,127 citations)
- DetectoRS: Detecting Objects With Recursive Feature Pyramid and Switchable Atrous Convolution (1,093 citations)
- Dynamic Head: Unifying Object Detection Heads With Attentions (1,013 citations)
- Multi-Attentional Deepfake Detection (942 citations)
- ReDet: A Rotation-Equivariant Detector for Aerial Object Detection (917 citations)
- You Only Look One-Level Feature (792 citations)
and 98 more in the interactive map.
Open-set and few-shot detection · 94 papers
- Simple Copy-Paste Is a Strong Data Augmentation Method for Instance Segmentation (1,154 citations)
- VinVL: Revisiting Visual Representations in Vision-Language Models (894 citations)
- Simultaneously Localize, Segment and Rank the Camouflaged Objects (571 citations)
- FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding (553 citations)
- Towards Open World Object Detection (542 citations)
- Adaptive Prototype Learning and Allocation for Few-Shot Segmentation (469 citations)
- Unbiased Mean Teacher for Cross-Domain Object Detection (374 citations)
- Anomaly Detection in Video via Self-Supervised and Multi-Task Learning (355 citations)
- VirTex: Learning Visual Representations From Textual Annotations (355 citations)
- Distribution Alignment: A Unified Framework for Long-Tail Visual Recognition (300 citations)
- HOTR: End-to-End Human-Object Interaction Detection With Transformers (291 citations)
- Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection (244 citations)
and 82 more in the interactive map.
Weakly supervised detection and segmentation · 32 papers
- Open-Vocabulary Object Detection Using Captions (417 citations)
- BoxInst: High-Performance Instance Segmentation With Box Annotations (331 citations)
- Anti-Adversarially Manipulated Attributions for Weakly and Semi-Supervised Semantic Segmentation (274 citations)
- Railroad Is Not a Train: Saliency As Pseudo-Pixel Supervision for Weakly Supervised Semantic Segmentation (266 citations)
- Semi-Supervised Semantic Segmentation With Directional Context-Aware Consistency (245 citations)
- DatasetGAN: Efficient Labeled Data Factory With Minimal Human Effort (238 citations)
- Patch2Pix: Epipolar-Guided Pixel-Level Correspondences (228 citations)
- Instant-Teaching: An End-to-End Semi-Supervised Object Detection Framework (222 citations)
- BBAM: Bounding Box Attribution Map for Weakly Supervised Semantic and Instance Segmentation (218 citations)
- Non-Salient Region Object Mining for Weakly Supervised Semantic Segmentation (211 citations)
- Humble Teachers Teach Better Students for Semi-Supervised Object Detection (197 citations)
- Embedded Discriminative Attention Mechanism for Weakly Supervised Semantic Segmentation (163 citations)
and 20 more in the interactive map.
Tracking · 28 papers
- Transformer Meets Tracker: Exploiting Temporal Context for Robust Visual Tracking (866 citations)
- Graph Attention Tracking (495 citations)
- Quasi-Dense Similarity Learning for Multiple Object Tracking (459 citations)
- Track To Detect and Segment: An Online Multi-Object Tracker (401 citations)
- STMTrack: Template-Free Visual Tracking With Space-Time Memory Networks (359 citations)
- Towards More Flexible and Accurate Object Tracking With Natural Language: Algorithms and Benchmark (281 citations)
- Alpha-Refine: Boosting Tracking Performance by Precise Bounding Box Estimation (231 citations)
- Multiple Object Tracking With Correlation Learning (198 citations)
- SiamMOT: Siamese Multi-Object Tracking (153 citations)
- Detection, Tracking, and Counting Meets Drones in Crowds: A Benchmark (137 citations)
- Learning To Filter: Siamese Relation Network for Robust Tracking (136 citations)
- Improving Multiple Pedestrian Tracking by Track Management and Occlusion Handling (133 citations)
and 16 more in the interactive map.
3D detection and driving · 25 papers
- Center-Based 3D Object Detection and Tracking (2,039 citations)
- Multi-Modal Fusion Transformer for End-to-End Autonomous Driving (578 citations)
- Categorical Depth Distribution Network for Monocular 3D Object Detection (563 citations)
- Objects Are Different: Flexible Monocular 3D Object Detection (309 citations)
- (AF)2-S3Net: Attentive Feature Fusion With Adaptive Feature Selection for Sparse Semantic Segmentation Network (264 citations)
- Delving Into Localization Errors for Monocular 3D Object Detection (262 citations)
- Robust Multimodal Vehicle Detection in Foggy Weather Using Complementary Lidar and Radar Signals (199 citations)
- ST3D: Self-Training for Unsupervised Domain Adaptation on 3D Object Detection (194 citations)
- RSN: Range Sparse Net for Efficient, Accurate LiDAR 3D Object Detection (164 citations)
- VIP-DeepLab: Learning Visual Perception With Depth-Aware Video Panoptic Segmentation (145 citations)
- Depth-Conditioned Dynamic Message Propagation for Monocular 3D Object Detection (141 citations)
- Objectron: A Large Scale Dataset of Object-Centric Videos in the Wild With Pose Annotations (140 citations)
and 13 more in the interactive map.
