GE-AdvGAN: Improving the transferability of adversarial samples by gradient editing-based adversarial generative model.
Zhiyu ZhuHuaming ChenXinyi WangJiayu ZhangZhibo JinKim-Kwang Raymond ChooJun ShenDong YuanPublished in: SDM (2024)
Keyphrases
- generative model
- probabilistic model
- mixture model
- bayesian framework
- data sets
- prior knowledge
- em algorithm
- discriminative learning
- object categories
- posterior probability
- semi supervised
- training set
- fully bayesian
- latent dirichlet allocation
- topic models
- expectation maximization
- markov chain
- maximum likelihood
- generative and discriminative models