RID-Noise: Towards Robust Inverse Design under Noisy Environments (2021-12-07T00:00:00.000000Z)

TL;DR

This work proposes Robust Inverse Design under Noise (RID-Noise), which can utilize existing data to train a conditional invertible neural network, and estimates the robustness of a design parameter by its predictability, measured by the prediction error of a forward neural network.

Abstract

From an engineering perspective, a design should not only perform well in an ideal condition, but should also resist noises. Such a design methodology, namely robust design, has been widely implemented in the industry for product quality control. However, classic robust design requires a lot of evaluations for a single design target, while the results of these evaluations could not be reused for a new target. To achieve data-efficient robust design, we propose Robust Inverse Design under Noise (RID-Noise), which can utilize existing data to train a conditional invertible neural network. Specifically, we estimate the robustness of a design parameter by its predictability, measured by the prediction error of a forward neural network. We also define a sample-wise weight, which can be used in the maximum weighted likelihood estimation of an inverse model based on a conditional invertible neural network. With the visual results from experiments, we clearly justify how RID-Noise works by learning the distribution and robustness from data. Further experiments on several real-world benchmark tasks with noises confirm that our method is more effective than other state-of-the-art inverse design methods. Code and supplementary is publicly available at https://github.com/ThyrixYang/rid-noise-aaai22

Authors

De-chuan Zhan

6 papers

Hao Ma

2 papers

Jia-Qi Yang

1 papers

TL;DR

Abstract

Authors

References30 items

Conditional Invertible Neural Networks for Diverse Image-to-Image Translation

Benchmarking Invertible Architectures on Inverse Problems

Neural-adjoint method for the inverse design of all-dielectric metasurfaces.

Deep learning for the design of photonic structures

Benchmarking deep inverse models over time, and the neural-adjoint method

Deep neural networks for the evaluation and design of photonic devices

Machine learning and Design of Experiments: Alternative approaches or complementary methodologies for quality improvement?

Deep learning enabled inverse design in nanophotonics

Normalizing Flows for Probabilistic Modeling and Inference

Deep learning for accelerated all-dielectric metasurface design.

Conditioning by adaptive sampling for robust design

Analyzing Inverse Problems with Invertible Neural Networks

Glow: Generative Flow with Invertible 1x1 Convolutions

Noise2Noise: Learning Image Restoration without Clean Data

Optimizing Chemical Reactions with Deep Reinforcement Learning

Machine learning under the spotlight

Training Deep Neural Networks for the Inverse Design of Nanophotonic Structures

Density estimation using Real NVP

Learning Structured Output Representation using Deep Conditional Generative Models

Conditional Generative Adversarial Nets

Auto-Encoding Variational Bayes

Weighted likelihood in Bayesian inference

Robust likelihood functions in Bayesian inference

Principles of robust design methodology

Pattern Recognition and Machine Learning

Local Likelihood Estimation

Robust Design: An Overview

Design for latitude using TOPT

Introduction to off-line quality control

Free-form diffractive metagrating design based on generative adversarial networks

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names