Paper ID: 2206.09027

Landscape Learning for Neural Network Inversion

Ruoshi Liu, Chengzhi Mao, Purva Tendulkar, Hao Wang, Carl Vondrick

Many machine learning methods operate by inverting a neural network at inference time, which has become a popular technique for solving inverse problems in computer vision, robotics, and graphics. However, these methods often involve gradient descent through a highly non-convex loss landscape, causing the optimization process to be unstable and slow. We introduce a method that learns a loss landscape where gradient descent is efficient, bringing massive improvement and acceleration to the inversion process. We demonstrate this advantage on a number of methods for both generative and discriminative tasks, including GAN inversion, adversarial defense, and 3D human pose reconstruction.

Submitted: Jun 17, 2022