Fetching the paper…

Model Inversion Attack against Transfer Learning: Inverting a Model without Accessing It · Around