Conference on Computer Vision and Pattern Recognition
Fraction of papers covering each topic over time
Topics over-represented at CVPR vs. the field average (2025)
| Year | Title | Citations | Links |
|---|---|---|---|
| 2021 | High-Resolution Image Synthesis with Latent Diffusion Models | 21,909 | S2 · arXiv |
| 2019 | Momentum Contrast for Unsupervised Visual Representation Learning | 14,254 | S2 · arXiv |
| 2021 | Masked Autoencoders Are Scalable Vision Learners | 10,415 | S2 · arXiv |
| 2022 | YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors | 9,437 | S2 · arXiv |
| 2019 | nuScenes: A Multimodal Dataset for Autonomous Driving | 7,395 | S2 · arXiv |
| 2022 | A ConvNet for the 2020s | 7,304 | S2 · arXiv |
| 2019 | Analyzing and Improving the Image Quality of StyleGAN | 6,705 | S2 · arXiv |
| 2019 | EfficientDet: Scalable and Efficient Object Detection | 6,432 | S2 · arXiv |
| 2019 | ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks | 5,401 | S2 · arXiv |
| 2019 | Generalized Intersection Over Union: A Metric and a Loss for Bounding Box Regression | 5,258 | S2 · arXiv |
| 2019 | Deep High-Resolution Representation Learning for Human Pose Estimation | 4,836 | S2 · arXiv |
| 2020 | Exploring Simple Siamese Representation Learning | 4,751 | S2 · arXiv |
| 2023 | Improved Baselines with Visual Instruction Tuning | 4,381 | S2 · arXiv |
| 2021 | Coordinate Attention for Efficient Mobile Network Design | 4,368 | S2 · arXiv |
| 2019 | DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation | 4,290 | S2 · arXiv |
| 2020 | Taming Transformers for High-Resolution Image Synthesis | 3,891 | S2 · arXiv |
| 2022 | DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation | 3,850 | S2 · arXiv |
| 2019 | Scalability in Perception for Autonomous Driving: Waymo Open Dataset | 3,701 | S2 · arXiv |
| 2019 | GhostNet: More Features From Cheap Operations | 3,684 | S2 · arXiv |
| 2020 | Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformers | 3,464 | S2 · arXiv |