2017年GAN 計算機視覺相關paper匯總

主要收集從 2016年11月(CVPR2017 deadline)到現在的 生成對抗網路(GAN)相關paper (按arXiv發表順序), 有遺漏歡迎補充。

  1. [1611.04076] Least Squares Generative Adversarial Networks (Cycle GAN的D用了其中的方法,將Loss改為L2 Loss,訓練穩定性提高了,好於傳統的cross-entropy loss)
  2. [1611.07004] Image-to-Image Translation with Conditional Adversarial Networks (pix2pix) 這篇中70x70的patchD在cyclegan中也有延續。 註:這篇中是pair來訓練的
  3. [1612.05363] Learning Residual Images for Face Attribute Manipulation (CVPR 2017) 這一篇比較早就用dual learning了,比cycle gan都早了3個月啊。但是他用的cycle loss是D的Loss,而不是pixel level的L1 Loss。
  4. [1612.07828] Learning from Simulated and Unsupervised Images through Adversarial Training (CVPR 2017)
  5. [1701.07717] Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro (ICCV2017) 這篇索性拿G產生的圖片來做數據增強,做了一個半監督學習框架,也容易理解。訓傳統分類ResNet,而沒有使用D。 因為D現在還是比較淺?在Fine-grain dataset上也有提高。
  6. [1701.02676] Unsupervised Image-to-Image Translation with Generative Adversarial Networks 之前我們可以用cgan來指定生成什麼domain。作者多加了一個分類器來預測隨機input z(在學習映射完成後,z其實是有semantic含義的)所以如果除了c以外,我們還能指定z。
  7. [1701.07875] Wasserstein GAN
  8. [1703.02291] Triple Generative Adversarial Nets
  9. [1703.05192] DiscoGAN: Learning to Discover Cross-Domain Relations with Generative Adversarial Networks (ICML2017) 同一個世界,同一個夢想 喜歡這篇裡面的圖和實驗。
  10. [1703.10593] CycleGAN: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks (ICCV2017) 同一個世界,同一個夢想 代碼很solid。 D用的是L2 Loss, Cycle Loss和Identity Loss都是L1 Loss。
  11. [1704.02510] DualGAN:Unsupervised Dual Learning for Image-to-Image Translation (ICCV2017) 同一個世界,同一個夢想
  12. [1704.00028] Improved Training of Wasserstein GANs
  13. Visual Saliency Prediction with Generative Adversarial Networks 阿岳感覺很一般
  14. Boosting Generative Models
  15. Towards Realistic High-Resolution Image Blending
  16. [1704.05838] Generative Face Completion
  17. [1704.04086] Global and Local Perception GAN for Photorealistic and Identity Preserving Frontal View Synthesis (ICCV 2017) 人臉pose旋轉,需要pair來訓練。一個網路做global的人臉,一個網路切5個關鍵點。最後融合到一起。
  18. [1704.04131] Neural Face Editing with Intrinsic Image Disentangling (CVPR2017)
  19. [1706.05274v2] Perceptual Generative Adversarial Networks for Small Object Detection
  20. Decomposing Motion and Content for Video Generation
  21. [1706.07068]Creative Adversarial Networks, Generating "Art" by Learning About Styles and Deviating from Style Norms 生成藝術作品
  22. [1707.03124] Adversarial Generation of Training Examples for Vehicle License Plate Recognition 用GAN來生成車牌,做數據增強

我的其他文章:

  • 用GAN生成的圖像做訓練?Yes!
  • 2017 ICCV 對抗生成網路GAN接收論文
  • 2017 ICCV 行人檢索/重識別 接受論文匯總

推薦閱讀:

港中大劉雲輝教授:自動駕駛、醫療手術、人機交互,機器視覺的應用潛力比你想像的要大
我國的車牌識別系統發展到了什麼水平?
從高層離職的Magic Leap談計算機視覺
[目標檢測] RON-Reverse Connection with Objectness Prior Networks for Object Detection
初見相關濾波與OTB

TAG:生成对抗网络GAN | 深度学习DeepLearning | 计算机视觉 |