Research map / 3D vision / Face and shape
HoloGAN: Unsupervised Learning of 3D Representations From Natural Images
Thu Nguyen-Phuoc, Chuan Li, Lucas Theis, Christian Richardt, Yong-Liang Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| apply random 3D rigid-body transformations during training (ablation) | HoloGAN · object datasets | pose control / disentanglement | Improved |
| use learnt constant tensor with z as AdaIN style instead of feeding z to first layer (ablation) | HoloGAN · object datasets | pose-identity disentanglement | Improved |
| add identity regulariser for 128x128 generation (ablation) | HoloGAN · 128x128 images | image quality / pose variation | Improved |
| use HoloGAN instead of WGAN-GP | WGAN-GP · Chairs 64x64 | KID (mean x100): 9.43 ± 0.24 -> 1.54 ± 0.07 | Improved |
| use HoloGAN instead of WGAN-GP | WGAN-GP · Cars 64x64 | KID (mean x100): 15.57 ± 0.29 -> 2.16 ± 0.09 | Improved |
| use HoloGAN instead of WGAN-GP | WGAN-GP · CelebA 64x64 | KID (mean x100): 1.63 ± 0.09 -> 2.87 ± 0.09 | Hurt |
| use HoloGAN instead of DCGAN | DCGAN · CelebA 64x64 | KID (mean x100): 1.81 ± 0.09 -> 2.87 ± 0.09 | Hurt |
| use HoloGAN instead of DCGAN | DCGAN · Cars 64x64 | KID (mean x100): 4.78 ± 0.11 -> 2.16 ± 0.09 | Improved |
Closest papers by meaning
- Unsupervised Learning of Efficient Geometry-Aware Neural Articulated Representations (ECCV 2022 · similarity 0.91)
- Inferring 3D Shapes from Image Collections Using Adversarial Networks (IJCV 2020 · similarity 0.90)
- Towards Unsupervised Learning of Generative Models for 3D Controllable Image Synthesis (CVPR 2020 · similarity 0.89)
- Learning Generative Models of Textured 3D Meshes From Real-World Images (ICCV 2021 · similarity 0.89)
- RGBD-GAN: Unsupervised 3D Representation Learning From Natural Image Datasets via RGBD Image Synthesis (ICLR 2020 · similarity 0.89)
- Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling (NeurIPS 2016 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
