CVPR
Conference on Computer Vision and Pattern Recognition
data from OpenAlex · report generated 2026-11-01Data details
Reporting period: 2016–2025. Data fetched: 2026-11-01 12:06 UTC
The abstract percentage is the share of indexed papers with abstracts, not the completeness of conference proceedings.
Coverage by year
| Year | Papers | With abstract |
|---|---|---|
| 2016 | 649 | 99.4% |
| 2017 | 791 | 99.4% |
| 2018 | 985 | 99.5% |
| 2019 | 1303 | 99.6% |
| 2020 | 1473 | 99.4% |
| 2021 | 1662 | 99.6% |
| 2022 | 2075 | 99.9% |
| 2023 | 2355 | 100.0% |
| 2024 | 2716 | 99.7% |
| 2025 | 2868 | 99.8% |
At a glance
- OpenAlex contains 16,877 papers for Conference on Computer Vision and Pattern Recognition between 2016 and 2025; output growing about 18% a year, from 649 to 2,868 papers.
- The fastest-growing phrase is “diffusion models”, up from 0.0% to 3.3% of papers.
- Shanghai Artificial Intelligence Laboratory gained the most share among institutions (0.0% → 2.6%); China among countries (34% → 50%).
- New vocabulary: diffusion models, vision-language, text-to-image, pre-training.
How the conversation changed
Share of papers matching the counting rule: the phrase appears in the title or more than once in the abstract. One phrase per family — variants such as “wafer hybrid bonding” are listed under it. Ranked by the gain between the first half of the period (2016–2020) and the second (2021–2025).
What is rising
0.0% → 3.3%
also
0.1% → 2.7%
also
1.3% → 3.1%
0.4% → 2.0%
What is fading
A falling share measures a change in vocabulary, not necessarily a decline in the technique’s use.
7.8% → 2.7%
also , , ,
2.1% → 0.7%
1.7% → 0.5%
also
1.3% → 0.1%
also
New vocabulary
Phrases that were almost absent before 2021 and now appear in at least one paper in a hundred. ✚ marks the first year a phrase shows up.
Phrase details
OpenAlex · paper-trends
How the field's topics shifted
“visual / text / vision-language” grew from 7.3% to 16% of papers, while “deep / network / neural networks” fell from 25% to 11%.
OpenAlex · paper-trends
- — 3d, view, pose, reconstruction, scene, estimation (3482 papers)
- — networks, neural, deep, network, neural networks, convolutional (2611 papers)
- — language, visual, vision, models, text, vision-language (2290 papers)
- — video, temporal, videos, motion, action, frames (2023 papers)
- — diffusion, generation, diffusion models, text, quality, models (1809 papers)
Show all topics
- — object, detection, object detection, objects, detectors, detector (1623 papers)
- — domain, adaptation, target, source, domain adaptation, domains (1530 papers)
- — segmentation, supervised, semantic, semantic segmentation, labels, self (1505 papers)
Who publishes
Papers with at least one author at the institution (a paper counts once per institution; subsidiaries and institutes are credited to their parent). The arrow is the change in share of papers between the two halves of the period.
Institution details
OpenAlex · paper-trends