An official website of the United States government
Generative Adversarial Networks (GANs) have shown remarkable success in various generative design tasks, from topology optimization to material design, and shape parametrization. However, most generative design approaches based on GANs lack evaluation mechanisms to ensure the generation of diverse samples. In addition, no GAN-based generative design model incorporates user sentiments in the loss function to generate samples with high desirability from the aggregate perspectives of users. Motivated by these knowledge gaps, this paper builds and validates a novel GAN-based generative design model with an offline design evaluation function to generate samples that are not only realistic, but also diverse and desirable. A multimodal Data-driven Design Evaluation (DDE) model is developed to guide the generative process by automatically predicting user sentiments for the generated samples based on large-scale user reviews of previous designs. This paper incorporates DDE into the StyleGAN structure, a state-of-the-art GAN model, to enable data-driven generative processes that are innovative and user-centered. The results of experiments conducted on a large dataset of footwear products demonstrate the effectiveness of the proposed DDE-GAN in generating high-quality, diverse, and desirable concepts.
more »
« less
Realizing GANs via a Tunable Loss Function
We introduce a tunable GAN, called α-GAN, parameterized by α∈(0, ∞], which interpolates between various f-GANs and Integral Probability Metric based GANs (under constrained discriminator set). We construct α− GAN using a supervised loss function, namely, α− loss, which is a tunable loss function capturing several canonical losses. We show that α− GAN is intimately related to the Arimoto divergence, which was first proposed by Österriecher (1996), and later studied by Liese and Vajda (2006). We posit that the holistic understanding that α− GAN introduces will have practical benefits of addressing both the issues of vanishing gradients and mode collapses.
more »
« less
- PAR ID:
- 10331824
- Date Published:
- Journal Name:
- 2021 IEEE Information Theory Workshop (ITW)
- Page Range / eLocation ID:
- 1 to 6
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
A Generative Adversarial Network (GAN) is an unsupervised generative framework to generate a sample distribution that is identical to the data distribution. Recently, mix strategy multi-generator/discriminator GANs have been shown to outperform single pair GANs. However, the mixed model suffers from the problem of linearly growing training time. Also, imbalanced training among generators makes it difficult to parallelize. In this paper, we propose a balanced mix-generator GAN that works in parallel by mixing multiple disjoint generators to approximate the real distribution. The weights of the discriminator and the classifier are controlled by a balance strategy. We also present an efficient loss function, to force each generator to embrace few modes with a high probability. Our model is naturally adaptive to large parallel computation frameworks. Each generator can be trained on multiple GPUs asynchronously. We have performed extensive experiments on synthetic datasets, MNIST1000, CIFAR-10, and ImageNet. The results establish that our model can achieve the state-of-the-art performance (in terms of the modes coverage and the inception score), with significantly reduced training time. We also show that the missing mode problem can be relieved with a growing number of generators.more » « less
-
Abstract Generative adversarial networks (GAN) have witnessed tremendous growth in recent years, demonstrating wide applicability in many domains. However, GANs remain notoriously difficult for people to interpret, particularly for modern GANs capable of generating photo‐realistic imagery. In this work we contribute a visual analytics approach for GAN interpretability, where we focus on the analysis and visualization of GAN disentanglement. Disentanglement is concerned with the ability to control content produced by a GAN along a small number of distinct, yet semantic, factors of variation. The goal of our approach is to shed insight on GAN disentanglement, above and beyond coarse summaries, instead permitting a deeper analysis of the data distribution modeled by a GAN. Our visualization allows one to assess a single factor of variation in terms of groupings and trends in the data distribution, where our analysis seeks to relate the learned representation space of GANs with attribute‐based semantic scoring of images produced by GANs. Through use‐cases, we show that our visualization is effective in assessing disentanglement, allowing one to quickly recognize a factor of variation and its overall quality. In addition, we show how our approach can highlight potential dataset biases learned by GANs.more » « less
-
null (Ed.)Generative adversarial networks (GANs) have been recently proposed for the synthesis of RF micro-Doppler signatures to mitigate the problem of low sample support and enable the training of deeper neural networks (DNNs) for improved RF signal classification. However, when applied to human micro-Doppler signatures for gait analysis, GANs suffer from systemic kinematic discrepancies that degrade performance. As a solution to this problem, this paper proposes the design of a physics-aware loss function and multi-branch GAN architecture. Our results show that RF gait signatures synthesized using the proposed approached have greater correlation and similarity to measured RF gait signatures, while also improving the accuracy in classifying five different gaits.more » « less
-
null (Ed.)Maximal α-leakage is a tunable measure of information leakage based on the quality of an adversary's belief about an arbitrary function of private data based on public data. The parameter α determines the loss function used to measure the quality of a belief, ranging from log-loss at α = 1 to the probability of error at α = ∞. We review its definition and main properties, including extensions to α <; 1, robustness to side information, and relationship to Rényi differential privacy.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/ITW48936.2021.9611499
