GEAC: Generating and Evaluating Handwritten Arabic Characters Using Generative Adversarial Networks
| Autor: | Norah Alqahtani, Sreela Sasi, Aya Bashy, Nahed Ali, Nujood Almajnooni, Tarik Alafif, Aisha Fallatah, Telawa Albarakati, Ahad Allhabi, Tahani Alkhodidi, Lamia Aljoudi |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Arabic characters
Computer science Research areas business.industry Value (computer science) Computational intelligence Pattern recognition 02 engineering and technology Image synthesis Adversarial system 020204 information systems 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business Generative adversarial network Generative grammar |
| Zdroj: | 2021 International Conference on Computational Intelligence and Knowledge Economy (ICCIKE). |
| DOI: | 10.1109/iccike51210.2021.9410746 |
| Popis: | Generative Adversarial Network (GAN) has made a breakthrough and great success in many research areas in computer vision. Different GANs generate different outputs. In this research work, we apply different GANs to generate handwritten Arabic characters. A basic GAN, Vanilla GAN, Deep Convolutional GAN (DCGAN), Bidirectional GAN (BiGAN), and Wasserstein GAN (WGAN) are used. Then, the results of the generated images are evaluated using native-Arabic human and Frechet Inception Distance (FID). The qualitative and quantitative results are provided for the images generation and evaluation. In experimental evaluation, WGAN achieves better results in FID with a value of 96.007. On the other hand, DCGAN achieves better results in native-Arabic human evaluation with a value of 35%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|