Generative Adversarial Network(GAN) Papers with 1k+ citations

Updated on Feb. 2021
To understand the direction of GAN research, I used Microsoft Academic to gather the papers with 1k+ citations with the keyword “Generative Adversarial”.
Except for original GAN paper and Deep convolutional GAN, conditional GAN paper, the statistic shows that most of the popular articles are about unpaired data generation, handling optimization issues, higher resolution, and embedding conditions.
Other worth mentioning topics are about text to image transformation, domain adaptation, RL, and 3D object synthesis.
Keep updating: https://hackmd.io/2CD_XoX6SbmGadAq5tVRmw
Papers sorted with #citations
- (Vanilla GAN)Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., … & Bengio, Y. (2014). Generative adversarial networks. arXiv preprint arXiv:1406.2661. Cited by 27497, 1st GAN paper
- (DCGAN)Radford, A., Metz, L., & Chintala, S. (2015). Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434. Cited by 8447, improve CNN in GAN using striding
- (Pix2Pix)Isola, P., Zhu, J. Y., Zhou, T., & Efros, A. A. (2017). Image-to-image translation with conditional adversarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1125–1134). Cited by 7899, Nvidia change noise-to-picture to picture-to-picture
- (cycleGAN)Zhu, J. Y., Park, T., Isola, P., & Efros, A. A. (2017). Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE international conference on computer vision (pp. 2223–2232). Cited by 7325, most well-known cycle loss with pix2pix discrimination
- (WGAN)Arjovsky, M., Chintala, S., & Bottou, L. (2017, July). Wasserstein generative adversarial networks. In International conference on machine learning (pp. 214–223). PMLR. Cited by 6202, replace JS divergence with Wasserstein distance
- (SRGAN)Ledig, C., Theis, L., Huszár, F., Caballero, J., Cunningham, A., Acosta, A., … & Shi, W. (2017). Photo-realistic single image super-resolution using a generative adversarial network. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4681–4690). Cited by 5026, the most well-known super-resolution paper
- (conditional GAN)Mirza, M., & Osindero, S. (2014). Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784. Cited by 4729, 1st conditional GAN
- (WGAN-GP)Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V., & Courville, A. (2017). Improved training of wasserstein gans. arXiv preprint arXiv:1704.00028.Cited by 4187, accelerated WGAN
- (DANN)Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., … & Lempitsky, V. (2016). Domain-adversarial training of neural networks. The journal of machine learning research, 17(1), 2096–2030., Cited by 2737 Domain Adaptation
- (ProGAN)Karras, T., Aila, T., Laine, S., & Lehtinen, J. (2017). Progressive growing of gans for improved quality, stability, and variation. arXiv preprint arXiv:1710.10196., Cited by 2597, layer by layer learning improve quality
- (InfoGAN)Chen, X., Duan, Y., Houthooft, R., Schulman, J., Sutskever, I., & Abbeel, P. (2016). Infogan: Interpretable representation learning by information maximizing generative adversarial nets. arXiv preprint arXiv:1606.03657. Cited by 2522, adding conditional variables into input noise for conditional GAN
- (2time scale update)Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., & Hochreiter, S. (2017). Gans trained by a two time-scale update rule converge to a local nash equilibrium. arXiv preprint arXiv:1706.08500. Cited by 2235, run discriminator more times to improve training
- (LSGAN)Mao, X., Li, Q., Xie, H., Lau, R. Y., Wang, Z., & Paul Smolley, S. (2017). Least squares generative adversarial networks. In Proceedings of the IEEE international conference on computer vision (pp. 2794–2802). Cited by 1991, replace Wasserstein distance with JS divergence and L2 error of a linear output
- (ADDA)Tzeng, E., Hoffman, J., Saenko, K., & Darrell, T. (2017). Adversarial discriminative domain adaptation. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 7167–7176). Cited by 1924, apply GAN on classifier domain transfer
- (GAN-CLS)Reed, S., Akata, Z., Yan, X., Logeswaran, L., Schiele, B., & Lee, H. (2016, June). Generative adversarial text to image synthesis. In International Conference on Machine Learning (pp. 1060–1069). PMLR.Cited by 1860, GAN text to image synthesis
- (LAPGAN)Denton, E., Chintala, S., Szlam, A., & Fergus, R. (2015). Deep generative image models using a laplacian pyramid of adversarial networks. arXiv preprint arXiv:1506.05751. Cited by 1791, learning self upsample
- (Spectral Normalization)Miyato, T., Kataoka, T., Koyama, M., & Yoshida, Y. (2018). Spectral normalization for generative adversarial networks. arXiv preprint arXiv:1802.05957. Cited by 1750, real 1-Lipschitz GAN loss
- (styleGAN)Karras, T., Laine, S., & Aila, T. (2019). A style-based generator architecture for generative adversarial networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 4401–4410). Cited by 1634, AdaIN style transfer with high quality
- (stackGAN)Zhang, H., Xu, T., Li, H., Zhang, S., Wang, X., Huang, X., & Metaxas, D. N. (2017). Stackgan: Text to photo-realistic image synthesis with stacked generative adversarial networks. In Proceedings of the IEEE international conference on computer vision (pp. 5907–5915).Cited by 1475, text to image 2 stage image improvement
- (BigGAN)Brock, A., Donahue, J., & Simonyan, K. (2018). Large scale GAN training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096. Cited by 1455, stabilize training with large batch + truncated sampling input noise
- (UNIT)Liu, M. Y., Breuel, T., & Kautz, J. (2017). Unsupervised image-to-image translation networks. arXiv preprint arXiv:1703.00848.Cited by 1441, common latent space for domain A and B
- (StarGAN)Choi, Y., Choi, M., Kim, M., Ha, J. W., Kim, S., & Choo, J. (2018). Stargan: Unified generative adversarial networks for multi-domain image-to-image translation. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 8789–8797).Cited by 1398, multiple cross-domain style transfer
- (MIXUP)Zhang, H., Cisse, M., Dauphin, Y. N., & Lopez-Paz, D. (2017). mixup: Beyond empirical risk minimization. arXiv preprint arXiv:1710.09412. Cited by 1389, change Empirical Risk Minimization to Vincinal Risk Minimization for data augmentation
- (SeqGAN)Yu, L., Zhang, W., Wang, J., & Yu, Y. (2017, February). Seqgan: Sequence generative adversarial nets with policy gradient. In Proceedings of the AAAI conference on artificial intelligence (Vol. 31, №1).Cited by 1330, RL sequence to sequence GAN
- (SAGANs)Zhang, H., Goodfellow, I., Metaxas, D., & Odena, A. (2019, May). Self-attention generative adversarial networks. In International conference on machine learning (pp. 7354–7363). PMLR.Cited by 1327, intra-image self-attention GAN
- (DiscoGAN)Kim, T., Cha, M., Kim, H., Lee, J. K., & Kim, J. (2017, July). Learning to discover cross-domain relations with generative adversarial networks. In International Conference on Machine Learning (pp. 1857–1865). PMLR.Cited by 1180, cycle GAN with single discrimination
- (Problem of JS divergence)Arjovsky, M., & Bottou, L. (2017). Towards principled methods for training generative adversarial networks. arXiv preprint arXiv:1701.04862.Cited by 1140, JS divergence cannot distinguish between those untouched conditions
- (GAIL)Ho, J., & Ermon, S. (2016). Generative adversarial imitation learning. arXiv preprint arXiv:1606.03476.Cited by 1083, GAN in imitation learning in RL
- (3D-GAN)Wu, J., Zhang, C., Xue, T., Freeman, W. T., & Tenenbaum, J. B. (2016). Learning a probabilistic latent space of object shapes via 3d generative-adversarial modeling. arXiv preprint arXiv:1610.07584.Cited by 1063, GAN 3d object
- (CyCADA)Hoffman, J., Tzeng, E., Park, T., Zhu, J. Y., Isola, P., Saenko, K., … & Darrell, T. (2018, July). Cycada: Cycle-consistent adversarial domain adaptation. In International conference on machine learning (pp. 1989–1998). PMLR. Cited by 1066, cycle GAN in domain adaptation
- (DualGAN)Yi, Z., Zhang, H., Tan, P., & Gong, M. (2017). Dualgan: Unsupervised dual learning for image-to-image translation. In Proceedings of the IEEE international conference on computer vision (pp. 2849–2857). Cited by 1014, cycle GAN with U-Net
Type of Papers (order by publish time)
Origian GAN type
2014 Vanilla GAN, 27497 cites
2014 conditional GAN, 4729 cites
2015 DCGAN, 8447 cites
Unpaired GAN
2017 cycleGAN, 7325 cites
2017 UNIT, 1441 cites
2017 DualGAN, 1014 cites
2017 DiscoGAN, 1180 cites
2018 StarGAN, 1398 cites
Higher resolution
2015 LAPGAN (generated resolution), 1791 cites
2016 SRGAN (upsamled resolution), 5026 cites
2017 ProGAN (generated resolution), 2597 cites
2018 SAGANs (generated resolution), 1327 cites
2019 styleGAN (generated resolution), 1634 cites
Better Embedding Condition
2016 InfoGAN, 2522 cites
2016 Pix2Pix, 7899 cites
2017 MIXUP, 1389 cites
2018 BigGAN, 1455 cites
Handeling optimization issues
2016 LSGAN, 1991 cites
2017 Problem of JS divergence, 1140 cites
2017 WGAN, 6202 cites
2017 WGAN-GP, 4187 cites
2017 2time scale update, 2235 cits
2018 Spectral Normalization, 1750 cites
Domain Adaptation
2016 DANN, 2737 cites
2017 ADDA, 1924 cites
2017 CyCADA, 1066 cites
Text to image
2016 GAN-CLS, 1860 cites
2017 stackGAN, 1475 cites
Text to Text
2017 SeqGAN, 1330 cites
RL
2016 GAIL, 1083 cites
3D object synthesis
2016 3D-GAN, 1063 cites