Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Controllable 3D Image Generation
This project attempts to learn object geometry and appearance from a set of 2D images and allows for scale specific controlling. We have also witnessed many great processes in realistic controllable 2D image synthesis [1] and pleasant 3D image results by tacking leverage the recent advance in volume rendering [2,3,4]. The core idea of this project is to extend the recent 3D generator that enables a level of control on both appearance and geometry [1,4].
Keywords: 3D generation; 3D representation
Reference:
[1] Karras et al. “A Style-Based Generator Architecture for Generative Adversarial Networks.” CVPR 2019
[2] Gas et al. "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned from Images." NeurIPS 2022
[3] Chan et al. "Efficient Geometry-aware 3D Generative Adversarial Networks." CVPR 2022
[4] Chen et al. "SofGAN: A Portrait Image Generator with Dynamic Styling." TOG
Reference:
[1] Karras et al. “A Style-Based Generator Architecture for Generative Adversarial Networks.” CVPR 2019
[2] Gas et al. "GET3D: A Generative Model of High Quality 3D Textured Shapes Learned from Images." NeurIPS 2022
[3] Chan et al. "Efficient Geometry-aware 3D Generative Adversarial Networks." CVPR 2022
[4] Chen et al. "SofGAN: A Portrait Image Generator with Dynamic Styling." TOG
Tasks:
- To familiarize the papers below and the code in method [2].
- Obtain some baseline results of the method [2] on different datasets.
- To analyze the performance and properties of the method [2].
- [Optional] To extend the framework with a controllable network design.
Requirement: 3D Vision; PyTorch; Deep Learning
Tasks:
- To familiarize the papers below and the code in method [2].
- Obtain some baseline results of the method [2] on different datasets.
- To analyze the performance and properties of the method [2].
- [Optional] To extend the framework with a controllable network design.