3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network
3D shape generation is widely applied in various industries to create, visualize, and analyse complex data, designs, and simulations. Typically, 3D shape generation uses a large dataset of 3D shapes as the input. This paper proposes a variational autoencoder with a signed distance function relativis...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/10/5925 |
| _version_ | 1827742247077019648 |
|---|---|
| author | Ebenezer Akinyemi Ajayi Kian Ming Lim Siew-Chin Chong Chin Poo Lee |
| author_facet | Ebenezer Akinyemi Ajayi Kian Ming Lim Siew-Chin Chong Chin Poo Lee |
| author_sort | Ebenezer Akinyemi Ajayi |
| collection | DOAJ |
| description | 3D shape generation is widely applied in various industries to create, visualize, and analyse complex data, designs, and simulations. Typically, 3D shape generation uses a large dataset of 3D shapes as the input. This paper proposes a variational autoencoder with a signed distance function relativistic average generative adversarial network, referred to as 3D-VAE-SDFRaGAN, for 3D shape generation from 2D input images. Both the generative adversarial network (GAN) and variational autoencoder (VAE) algorithms are typical algorithms used to generate realistic 3D shapes. However, it is very challenging to train a stable 3D shape generation model using VAE-GAN. This paper proposes an efficient approach to stabilize the training process of VAE-GAN to generate high-quality 3D shapes. A 3D mesh-based shape is first generated using a 3D signed distance function representation by feeding a single 2D image into a 3D-VAE-SDFRaGAN network. The signed distance function is used to maintain inside–outside information in the implicit surface representation. In addition, a relativistic average discriminator loss function is employed as the training loss function. The polygon mesh surfaces are then produced via the marching cubes algorithm. The proposed 3D-VAE-SDFRaGAN is evaluated with the ShapeNet dataset. The experimental results indicate a notable enhancement in the qualitative performance, as evidenced by the visual comparison of the generated samples, as well as the quantitative performance evaluation using the chamfer distance metric. The proposed approach achieves an average chamfer distance score of 0.578, demonstrating superior performance compared to existing state-of-the-art models. |
| first_indexed | 2024-03-11T03:58:19Z |
| format | Article |
| id | doaj.art-12cf1421ceea46f08014a5f2ce7ed0b6 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-11T03:58:19Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-12cf1421ceea46f08014a5f2ce7ed0b62023-11-18T00:17:50ZengMDPI AGApplied Sciences2076-34172023-05-011310592510.3390/app131059253D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial NetworkEbenezer Akinyemi Ajayi0Kian Ming Lim1Siew-Chin Chong2Chin Poo Lee3Faculty of Information Science and Technology, Multimedia University, Malacca 75450, MalaysiaFaculty of Information Science and Technology, Multimedia University, Malacca 75450, MalaysiaFaculty of Information Science and Technology, Multimedia University, Malacca 75450, MalaysiaFaculty of Information Science and Technology, Multimedia University, Malacca 75450, Malaysia3D shape generation is widely applied in various industries to create, visualize, and analyse complex data, designs, and simulations. Typically, 3D shape generation uses a large dataset of 3D shapes as the input. This paper proposes a variational autoencoder with a signed distance function relativistic average generative adversarial network, referred to as 3D-VAE-SDFRaGAN, for 3D shape generation from 2D input images. Both the generative adversarial network (GAN) and variational autoencoder (VAE) algorithms are typical algorithms used to generate realistic 3D shapes. However, it is very challenging to train a stable 3D shape generation model using VAE-GAN. This paper proposes an efficient approach to stabilize the training process of VAE-GAN to generate high-quality 3D shapes. A 3D mesh-based shape is first generated using a 3D signed distance function representation by feeding a single 2D image into a 3D-VAE-SDFRaGAN network. The signed distance function is used to maintain inside–outside information in the implicit surface representation. In addition, a relativistic average discriminator loss function is employed as the training loss function. The polygon mesh surfaces are then produced via the marching cubes algorithm. The proposed 3D-VAE-SDFRaGAN is evaluated with the ShapeNet dataset. The experimental results indicate a notable enhancement in the qualitative performance, as evidenced by the visual comparison of the generated samples, as well as the quantitative performance evaluation using the chamfer distance metric. The proposed approach achieves an average chamfer distance score of 0.578, demonstrating superior performance compared to existing state-of-the-art models.https://www.mdpi.com/2076-3417/13/10/59253D shape generationvariational autoencodergenerative adversarial networksigned distance functionrelativistic average |
| spellingShingle | Ebenezer Akinyemi Ajayi Kian Ming Lim Siew-Chin Chong Chin Poo Lee 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network Applied Sciences 3D shape generation variational autoencoder generative adversarial network signed distance function relativistic average |
| title | 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network |
| title_full | 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network |
| title_fullStr | 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network |
| title_full_unstemmed | 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network |
| title_short | 3D Shape Generation via Variational Autoencoder with Signed Distance Function Relativistic Average Generative Adversarial Network |
| title_sort | 3d shape generation via variational autoencoder with signed distance function relativistic average generative adversarial network |
| topic | 3D shape generation variational autoencoder generative adversarial network signed distance function relativistic average |
| url | https://www.mdpi.com/2076-3417/13/10/5925 |
| work_keys_str_mv | AT ebenezerakinyemiajayi 3dshapegenerationviavariationalautoencoderwithsigneddistancefunctionrelativisticaveragegenerativeadversarialnetwork AT kianminglim 3dshapegenerationviavariationalautoencoderwithsigneddistancefunctionrelativisticaveragegenerativeadversarialnetwork AT siewchinchong 3dshapegenerationviavariationalautoencoderwithsigneddistancefunctionrelativisticaveragegenerativeadversarialnetwork AT chinpoolee 3dshapegenerationviavariationalautoencoderwithsigneddistancefunctionrelativisticaveragegenerativeadversarialnetwork |