Landscape Learning for Neural Network Inversion

2022/06/17 by Liu, Ruoshi, Mao, Chengzhi, Tendulkar, Purva +2 · 4 citations
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2206.09027

Abstract

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.

Cited by

• TIE: A Training-Inversion-Exclusion Framework for Visually Interpretable and Uncertainty-Guided Out-of-Distribution Detection
• EXP-CAM: Explanation Generation and Circuit Discovery Using Classifier Activation Matching
• Activation Matching for Explanation Generation
• Investigating the Invertibility of Multimodal Latent Spaces: Limitations of Optimization-Based Methods

Related

• Regularized Training of Intermediate Layers for Generative Models for Inverse Problems
• Network Inversion of Convolutional Neural Nets
• Near Perfect GAN Inversion
• Network Inversion and Its Applications
• Inverting face embeddings with convolutional neural networks