RANGE-GAN: Design Synthesis Under Constraints Using Conditional Generative Adversarial Networks

2021 ◽  
pp. 1-16
Author(s):  
Amin Heyrani Nobari ◽  
Wei (Wayne) Chen ◽  
Faez Ahmed
Keyword(s):  
Generative Model ◽  
Inverse Design ◽  
Data Driven ◽  
Generative Adversarial Networks ◽  
Design Problems ◽  
Design Synthesis ◽  
Average Improvement ◽  
Multiple Alternative ◽  
Inverse Design Problem ◽  
Alternative Designs

Abstract Typical engineering design tasks require the effort to modify designs iteratively until they meet certain constraints, i.e., performance or attribute requirements. Past work has proposed ways to solve the inverse design problem, where desired designs are directly generated from specified requirements, thus avoid the trial and error process. Among those approaches, the conditional deep generative model shows great potential since 1) it works for complex high-dimensional designs and 2) it can generate multiple alternative designs given any condition. In this work, we propose a conditional deep generative model, Range-GAN, to achieve automatic design synthesis subject to range constraints. The proposed model addresses the sparse conditioning issue in data-driven inverse design problems by introducing a label-aware self-augmentation approach. We also propose a new uniformity loss to ensure generated designs evenly cover the given requirement range. Through a real-world example of constrained 3D shape generation, we show that the label-aware self-augmentation leads to an average improvement of14% on the constraint satisfaction for generated 3D shapes, and the uniformity loss leads to a 125% average increase on the uniformity of generated shapes' attributes. This work laid the foundation for data-driven inverse design problems where we consider range constraints and there are sparse regions in the condition space.

scholarly journals Range-GAN: Range-Constrained Generative Adversarial Network for Conditioned Design Synthesis

2021 ◽  
Author(s):  
Amin Heyrani Nobari ◽  
Wei Chen ◽  
Faez Ahmed
Keyword(s):  
Generative Model ◽  
Inverse Design ◽  
Data Driven ◽  
Generative Adversarial Network ◽  
Design Problems ◽  
Design Synthesis ◽  
Adversarial Network ◽  
Average Improvement ◽  
Inverse Design Problem ◽  
Alternative Designs

Abstract Typical engineering design tasks require the effort to modify designs iteratively until they meet certain constraints, i.e., performance or attribute requirements. Past work has proposed ways to solve the inverse design problem, where desired designs are directly generated from specified requirements, thus avoid the trial and error process. Among those approaches, the conditional deep generative model shows great potential since 1) it works for complex high-dimensional designs and 2) it can generate multiple alternative designs given any condition. In this work, we propose a conditional deep generative model, Range-GAN, to achieve automatic design synthesis subject to range constraints. The proposed model addresses the sparse conditioning issue in data-driven inverse design problems by introducing a label-aware self-augmentation approach. We also propose a new uniformity loss to ensure generated designs evenly cover the given requirement range. Through a real-world example of constrained 3D shape generation, we show that the label-aware self-augmentation leads to an average improvement of 14% on the constraint satisfaction for generated 3D shapes, and the uniformity loss leads to a 125% average increase on the uniformity of generated shapes’ attributes. This work laid the foundation for data-driven inverse design problems where we consider range constraints and there are sparse regions in the condition space. For further information and code for this paper please refer to http://decode.mit.edu/projects/rangegan/.

scholarly journals Predictive and generative machine learning models for photonic crystals

Nanophotonics ◽  
2020 ◽  
Vol 9 (13) ◽  
pp. 4183-4192 ◽  
Author(s):  
Thomas Christensen ◽  
Charlotte Loh ◽  
Stjepan Picek ◽  
Domagoj Jakobović ◽  
Li Jing ◽  
...  
Keyword(s):  
Machine Learning ◽  
Photonic Crystals ◽  
Optimization Techniques ◽  
Inverse Design ◽  
Data Driven ◽  
Generative Adversarial Networks ◽  
Learning Models ◽  
Test Bed ◽  
Data Set ◽  
Wide Range

AbstractThe prediction and design of photonic features have traditionally been guided by theory-driven computational methods, spanning a wide range of direct solvers and optimization techniques. Motivated by enormous advances in the field of machine learning, there has recently been a growing interest in developing complementary data-driven methods for photonics. Here, we demonstrate several predictive and generative data-driven approaches for the characterization and inverse design of photonic crystals. Concretely, we built a data set of 20,000 two-dimensional photonic crystal unit cells and their associated band structures, enabling the training of supervised learning models. Using these data set, we demonstrate a high-accuracy convolutional neural network for band structure prediction, with orders-of-magnitude speedup compared to conventional theory-driven solvers. Separately, we demonstrate an approach to high-throughput inverse design of photonic crystals via generative adversarial networks, with the design goal of substantial transverse-magnetic band gaps. Our work highlights photonic crystals as a natural application domain and test bed for the development of data-driven tools in photonics and the natural sciences.

scholarly journals Inverse design of glass structure with deep graph neural networks

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Qi Wang ◽  
Longfei Zhang
Keyword(s):  
Neural Networks ◽  
Monte Carlo ◽  
Configuration Space ◽  
Glass Structure ◽  
Stable State ◽  
Specific Property ◽  
Inverse Design ◽  
Data Driven ◽  
Common Belief ◽  
Graph Neural Networks

AbstractDirectly manipulating the atomic structure to achieve a specific property is a long pursuit in the field of materials. However, hindered by the disordered, non-prototypical glass structure and the complex interplay between structure and property, such inverse design is dauntingly hard for glasses. Here, combining two cutting-edge techniques, graph neural networks and swap Monte Carlo, we develop a data-driven, property-oriented inverse design route that managed to improve the plastic resistance of Cu-Zr metallic glasses in a controllable way. Swap Monte Carlo, as a sampler, effectively explores the glass landscape, and graph neural networks, with high regression accuracy in predicting the plastic resistance, serves as a decider to guide the search in configuration space. Via an unconventional strengthening mechanism, a geometrically ultra-stable yet energetically meta-stable state is unraveled, contrary to the common belief that the higher the energy, the lower the plastic resistance. This demonstrates a vast configuration space that can be easily overlooked by conventional atomistic simulations. The data-driven techniques, structural search methods and optimization algorithms consolidate to form a toolbox, paving a new way to the design of glassy materials.

Modeling and Inverse Design of Convective Heat Transfer in a Grooved Channel Using Proper Orthogonal Decomposition

2009 ◽  
Author(s):  
Hasan Gunes ◽  
Sertac Cadirci
Keyword(s):  
Heat Transfer ◽  
Heat Flux ◽  
Reynolds Number ◽  
Proper Orthogonal Decomposition ◽  
Convective Heat Transfer ◽  
Orthogonal Decomposition ◽  
Inverse Design ◽  
Heat Sources ◽  
Design Problems ◽  
Proper Orthogonal

In this study we show that the POD can be used as a useful tool to solve inverse design problems in thermo-fluids. In this respect, we consider a forced convection problem of air flow in a grooved channel with periodically mounted constant heat-flux heat sources. It represents a cooling problem in electronic equipments where the coolant is air. The cooling of electronic equipments with constant periodic heat sources is an important problem in the industry such that the maximum operating temperature must be kept below a value specified by the manufacturer. Geometric design in conjunction with the improved convective heat transfer characteristics is important to achieve an effective cooling. We obtain a model based on the proper orthogonal decomposition for the convection optimization problem such that for a given channel geometry and heat flux on the chip surface, we search for the minimum Reynolds number (i.e., inlet flow speed) for a specified maximum surface temperature. For a given geometry (l = 3.0 cm and h = 2.3 cm), we obtain a proper orthogonal decomposition (POD) model for the flow and heat transfer for Reynolds number in the range 1 and 230. It is shown that the POD model can accurately predict the flow and temperature field for off-design conditions and can be used effectively for inverse design problems.

scholarly journals Monte Carlo and Reconstruction Membership Inference Attacks against Generative Models

2019 ◽  
Vol 2019 (4) ◽  
pp. 232-249 ◽  
Author(s):  
Benjamin Hilprecht ◽  
Martin Härterich ◽  
Daniel Bernau
Keyword(s):  
Data Privacy ◽  
Information Leakage ◽  
Generative Models ◽  
Generative Model ◽  
Training Data ◽  
Generative Adversarial Networks ◽  
Data Sets ◽  
Success Rates ◽  
Model Quality ◽  
Type Formalization

Abstract We present two information leakage attacks that outperform previous work on membership inference against generative models. The first attack allows membership inference without assumptions on the type of the generative model. Contrary to previous evaluation metrics for generative models, like Kernel Density Estimation, it only considers samples of the model which are close to training data records. The second attack specifically targets Variational Autoencoders, achieving high membership inference accuracy. Furthermore, previous work mostly considers membership inference adversaries who perform single record membership inference. We argue for considering regulatory actors who perform set membership inference to identify the use of specific datasets for training. The attacks are evaluated on two generative model architectures, Generative Adversarial Networks (GANs) and Variational Autoen-coders (VAEs), trained on standard image datasets. Our results show that the two attacks yield success rates superior to previous work on most data sets while at the same time having only very mild assumptions. We envision the two attacks in combination with the membership inference attack type formalization as especially useful. For example, to enforce data privacy standards and automatically assessing model quality in machine learning as a service setups. In practice, our work motivates the use of GANs since they prove less vulnerable against information leakage attacks while producing detailed samples.

Sign in / Sign up

Export Citation Format

Share Document