CVPR2022-Paper-Statistics
Statistics and Visualization of main keyword of accepted papers for The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR 2022)
Inspired by CVPR-2021-Paper-Statistics
Wordcloud
Acceptance
| Year | Submissions | Acceptance | Acceptance rate |
|---|---|---|---|
| 2011 | 1677 | 438 | 26.10% |
| 2012 | 1933 | 466 | 24.10% |
| 2013 | 1798 | 472 | 26.20% |
| 2014 | 1807 | 540 | 29.90% |
| 2015 | 2123 | 602 | 28.40% |
| 2016 | 2145 | 643 | 29.90% |
| 2017 | 2680 | 783 | 29.20% |
| 2018 | 3359 | 979 | 29.10% |
| 2019 | 5160 | 1294 | 25.07% |
| 2020 | 6656 | 1468 | 22.13% |
| 2021 | 7015 | 1663 | 23.71% |
| 2022 | 8161 | 2067 | 25.32% |
Keyworks
- Most of the top keywords were maintained
- Image, Object, Detection, 3D, Video, Segmentation
-
Transfomer are about
$5\times$ as frequent, thanks for ViT- transfomer:
$35$ ->$177$ 😉
- transfomer:
Code Uasge
# 1. install packages
pip install matplotlib scipy pillow wordcloud nltk numpy
# 2. run script
python cvpr22.py --list misc/CVPR2022-papers.list