Generative-Discriminative Complementary Learning


  • Yanwu Xu University of Pittsburgh
  • Mingming Gong University of Pittsburgh
  • Junxiang Chen University of Pittsburgh
  • Tongliang Liu School of Computer Science
  • Kun Zhang Carnegie Mellon University
  • Kayhan Batmanghelich University of Pittsburgh



The majority of state-of-the-art deep learning methods are discriminative approaches, which model the conditional distribution of labels given inputs features. The success of such approaches heavily depends on high-quality labeled instances, which are not easy to obtain, especially as the number of candidate classes increases. In this paper, we study the complementary learning problem. Unlike ordinary labels, complementary labels are easy to obtain because an annotator only needs to provide a yes/no answer to a randomly chosen candidate class for each instance. We propose a generative-discriminative complementary learning method that estimates the ordinary labels by modeling both the conditional (discriminative) and instance (generative) distributions. Our method, we call Complementary Conditional GAN (CCGAN), improves the accuracy of predicting ordinary labels and is able to generate high-quality instances in spite of weak supervision. In addition to the extensive empirical studies, we also theoretically show that our model can retrieve the true conditional distribution from the complementarily-labeled data.




How to Cite

Xu, Y., Gong, M., Chen, J., Liu, T., Zhang, K., & Batmanghelich, K. (2020). Generative-Discriminative Complementary Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 34(04), 6526-6533.



AAAI Technical Track: Machine Learning