Object Design System by Interactive Evolutionary Computation Using GAN with Contour Images
| Autor: | Kaoru Arakawa, Chen Xin |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Training set
Basis (linear algebra) business.industry Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Process (computing) ComputerApplications_COMPUTERSINOTHERSYSTEMS Object design Interactive evolutionary computation Object (computer science) Image (mathematics) Computer vision Artificial intelligence Design objects business |
| Zdroj: | Human Centred Intelligent Systems ISBN: 9789811632631 KES-HCIS |
| Popis: | This paper proposes a method to design objects by interactive evolutionary computation (IEC) using generative adversarial network (GAN) with the contour image of the object which the user wants to design. Here, a conditional GAN is adopted in order to utilize the contour image. GANs can generate lifelike images of objects, however, it is hard to control the output, since GAN just generates images randomly with the distribution learned from training data. Thus, IEC is introduced here to control the latent vectors in the conditional GAN. Moreover, using the contour image, conditional GAN can efficiently generate images of objects which the user prefers. In the proposed method the latent vectors in the conditional GAN are optimized for each user, so that the design is satisfactory to him/her through the process of IEC. Since IEC is good at optimizing system parameters on the basis of human subjective criteria, the user can obtain preferable lifelike design image. Experiment results show the high performance of this method. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|