Science Explorer Back to mapMap
Discussion
Topic · Computer Vision and Pattern Recognition

Generative Adversarial Networks and Image Synthesis

Generative Adversarial Networks and Image Synthesis is a research topic within Computer Vision and Pattern Recognition. Science Explorer counts 19k research works in it since 1951. 9.7% of them reached the world's top 10% most cited for their field and year.

This cluster of papers focuses on the application of Generative Adversarial Networks (GANs) in image processing, including image synthesis, style transfer, representation learning, and unsupervised learning. The papers cover various techniques such as image inpainting, texture synthesis, and conditional generative models using deep learning and neural networks.

  • Generative Adversarial Networks
  • Image Synthesis
  • Deep Learning
  • Neural Networks
  • Image Inpainting
  • Style Transfer
  • Representation Learning
  • Unsupervised Learning
  • Conditional Generative Models
  • Texture Synthesis
Research works
19k
fractional, since 1951
In the world top 10%
1.8k
per year above
Top-10% rate
9.7%
share of its works in the world top 10%
Growth, 2018-21 → 2022-25
+118%
the tick is no change
Largest producer
China
50% of the 10 largest, 2022-now
Generative Adversarial Networks and Image Synthesis

Which countries lead Generative Adversarial Networks and Image Synthesis research?

By volume, China and the United States publish the most (4.9k and 1.4k works in 2022-now).

By volume, 2022-now

  1. 1 China 4.9k works
  2. 2 United States 1.4k works
  3. 3 India 1.4k works
  4. 4 South Korea 452 works
  5. 5 United Kingdom 379 works
  6. 6 Japan 319 works
  7. 7 Germany 313 works
  8. 8 France 209 works
  9. 9 Canada 203 works
  10. 10 Italy 200 works

How concentrated that is

The same countries as shares of everything the list above accounts for. A node where two countries do two thirds of the work and one spread evenly across twelve read alike as a ranking and not at all alike here.

China: 49.9%United States: 14.6%India: 14.4%South Korea: 4.6%6 others listed: 16.5%50%largest
China4,899 · 49.9%United States1,436 · 14.6%India1,416 · 14.4%South Korea452 · 4.6%6 others listed1,624 · 16.5%

Shares of the rows listed above, not of the whole node.

Which institutions lead Generative Adversarial Networks and Image Synthesis research?

By volume in 2022-now, Chinese Academy of Sciences publishes the most Generative Adversarial Networks and Image Synthesis research, followed by Shanghai Jiao Tong University and Tsinghua University.

By volume, 2022-now

  1. 1 Chinese Academy of SciencesChina 179 works
  2. 2 Shanghai Jiao Tong UniversityChina 108 works
  3. 3 Tsinghua UniversityChina 92 works
  4. 4 Zhejiang UniversityChina 87 works
  5. 5 University of Science and Technology of ChinaChina 86 works
  6. 6 Peking UniversityChina 69 works
  7. 7 Beijing University of Posts and TelecommunicationsChina 67 works
  8. 8 Sun Yat-sen UniversityChina 67 works
  9. 9 Centre National de la Recherche ScientifiqueFrance 64 works
  10. 10 Fudan UniversityChina 60 works

Who are the leading researchers in Generative Adversarial Networks and Image Synthesis?

The most-cited researchers publishing on Generative Adversarial Networks and Image Synthesis include M. Shoji, Kaiming He and Yoshua Bengio.

  1. 1 M. Shoji Japan 806k citations
  2. 2 Kaiming He Israel 562k citations
  3. 3 Yoshua Bengio Canada 476k citations
  4. 4 Geoffrey E. Hinton Canada 454k citations
  5. 5 Jun Wang China 340k citations
  6. 6 Xiangyu Zhang 327k citations
  7. 7 Trevor Hastie United States 301k citations
  8. 8 Jian Sun 281k citations
  9. 9 Xiaogang Wang China 260k citations
  10. 10 Yann LeCun United States 256k citations

Ranked by citations received across their whole record, among researchers with at least three works on this topic.

What are the most cited papers in Generative Adversarial Networks and Image Synthesis?

The most cited paper in Generative Adversarial Networks and Image Synthesis is "Deep learning" (Yann LeCun, Yoshua Bengio, Geoffrey E. Hinton, 2015), with 84,939 citations. Next come "Learning Multiple Layers of Features from Tiny Images" (25,341) and "Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks" (22,468).

Most cited papers
#PaperReleasedCitations
1 Deep learning Yann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature Open access 26 May 2015 84,939
2 Learning Multiple Layers of Features from Tiny Images Alex Krizhevsky Open access 2024 25,341
3 Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks Jun-Yan Zhu, Taesung Park, Phillip Isola et al. Conference paperOpen access Oct 2017 22,468
4 SSD: Single Shot MultiBox Detector Wei Liu, Dragomir Anguelov, Dumitru Erhan et al. · Lecture notes in computer science Conference paperOpen access 2016 21,849
5 GAN（Generative Adversarial Nets） 柴田 淳司 · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics Open access 15 Oct 2017 21,691
All 100 most cited papers, with filters and sorting

Where is Generative Adversarial Networks and Image Synthesis research done?

The largest centres of Generative Adversarial Networks and Image Synthesis research in 2022-now are Beijing (China), Shanghai (China), Xi'an (China) and Seoul (South Korea). Among places with at least 20 works in it, it is an unusually large share of all research in San Jose and Greater Noida.

Largest cities, 2022-now

  1. 1 BeijingChina 922 works
  2. 2 ShanghaiChina 395 works
  3. 3 Xi'anChina 250 works
  4. 4 SeoulSouth Korea 228 works
  5. 5 HangzhouChina 227 works
  6. 6 GuangzhouChina 226 works
  7. 7 NanjingChina 210 works
  8. 8 WuhanChina 202 works
  9. 9 ShenzhenChina 199 works
  10. 10 Hong KongChina 158 works

Where it is the local speciality

  1. San JoseUS · 37.9 works9.5×
  2. Greater NoidaIN · 38.9 works5.2×
← less than its size predicts1×more →

Location quotient: how much more of its research is in Generative Adversarial Networks and Image Synthesis than the world average.

See Generative Adversarial Networks and Image Synthesis on the map

Where is the best place to study Generative Adversarial Networks and Image Synthesis?

Among universities, judged by research, Nanyang Technological University, Tel Aviv University and University of Science and Technology of China score highest, combining excellence, specialisation, size, growth and international reach. Research strength is one signal when choosing where to study; it does not measure teaching.

0%20%40%mean 17.79%fractional works in this field (log) →share of work in the world top 10% →Tel Aviv UniversityNanyang Technologica…University of Hong K…University of Scienc…
above the meannear itbelow it

One dot per university in the first ten of the table below, at the default weights. The upper left is the interesting corner: small places doing unusually strong work.

Excellenceshare of its work in the world top 10%, 2022-now and all time averaged
40%
Specialisationhow much of its work is in this field, counted up to 10× the world share
10%
Sizeits output in this field, on a log scale
30%
Growthoutput growth, counted from -20% to +20%
10%
Internationalshare of its work with a co-author abroad
10%
Each slider sets a weight; the share it gets of the score is beside it.
#UniversityScoreTop 10%SpecialisationWorksGrowth
1Nanyang Technological UniversitySingapore 71.920.0%6.8×55 +153.0%
2Tel Aviv UniversityIsrael 68.635.7%2.3×19 +41.8%
3University of Science and Technology of ChinaChina 67.515.3%7.0×86 +129.7%
4University of Hong KongChina 67.319.2%3.9×37 +258.5%
5Tsinghua UniversityChina 66.616.6%3.9×92 +88.4%
6National University of SingaporeSingapore 65.318.7%3.5×43 +115.4%
7Hong Kong University of Science and TechnologyChina 62.414.6%8.0×34 +191.7%
8Zhejiang UniversityChina 61.713.9%3.1×87 +165.8%
9Shanghai Jiao Tong UniversityChina 61.410.4%3.6×108 +108.2%
10City University of Hong KongChina 60.513.5%5.5×32 +85.3%

Universities only. The score blends excellence (40%), size (30%), specialisation (10%), growth (10%), international reach (10%), 2022-now; growth compares 2018-21 with 2022-25 and counts up to +20%, since a jump from a small base is mostly noise.

Every institution in Generative Adversarial Networks and Image Synthesis, country by country

Is Generative Adversarial Networks and Image Synthesis research growing?

Output in 2022-2025 was 118% higher than in 2018-2021, peaking in 2025. The fastest-growing topics are Generative Adversarial Networks and Image Synthesis.

still counting19801990200020102020
grewheldshrank

The same series as a ribbon: one cell per year, darker for more. The line above answers how much; this answers when.

Its share of all research

Growing output can still be a shrinking share: all of science grows. This is Generative Adversarial Networks and Image Synthesis as a percentage of every work published that year, 0.035% in 2024.

still counting0.020%0.040%0.060%Generative Adversarial Networks and Image Synthesis0.060% in 202519801990200020102020

Fractional works of Generative Adversarial Networks and Image Synthesis over all fractional works that year. How the balance of science has shifted.

How much of it reaches the world top 10%

Year by year.

still counting5%10%15%Generative Adversarial Networks and Image Synthesis9.6% in 202520102020

Of each year's works in it, the share that reached the world's 10% most cited for their field and year. A field is the world for its own topics, so there is no other line to hold it against.

Which topics inside it are moving

Growth and decline on one axis around a shared zero. Two lists side by side hide the thing that matters: whether the growth dwarfs the decline, or the other way round.

Questions people ask

Who leads research on generative AI for images (GANs)?

By volume in 2022-now, Chinese Academy of Sciences publishes the most Generative Adversarial Networks and Image Synthesis research, followed by Shanghai Jiao Tong University and Tsinghua University. By volume, China and the United States publish the most (4.9k and 1.4k works in 2022-now).

All 1,273 questions in the Question Catalogue

Discussion

Ask a question or share a note about Generative Adversarial Networks and Image Synthesis. Posts are plain text and can be reported.