Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling

We study the problem of 3D object generation. We propose a novel framework, namely 3D Generative Adversarial Network (3D-GAN), which generates 3D objects from a probabilistic space by leveraging recent advances in volumetric convolutional networks and generative adversarial nets. The benefits of our model are three-fold: first, the use of an adversarial criterion, instead of traditional heuristic criteria, enables the generator to capture object structure implicitly and to synthesize high-quality 3D objects; second, the generator establishes a mapping from a low-dimensional probabilistic space to the space of 3D objects, so that we can sample objects without a reference image or CAD models, and explore the 3D object manifold; third, the adversarial discriminator provides a powerful 3D shape descriptor which, learned without supervision, has wide applications in 3D object recognition. Experiments demonstrate that our method generates high-quality 3D objects, and our unsupervisedly learned features achieve impressive performance on 3D object recognition, comparable with those of supervised learning methods.

PDF Abstract NeurIPS 2016 PDF NeurIPS 2016 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
3D Point Cloud Linear Classification ModelNet40 3D-GAN Overall Accuracy 83.3 # 9

Results from Other Papers


Task Dataset Model Metric Name Metric Value Rank Source Paper Compare
3D Shape Classification Pix3D 3D-VAE-GAN R@1 0.02 # 3
R@16 0.21 # 3
R@2 0.03 # 3
R@32 0.34 # 3
R@4 0.07 # 3
R@8 0.12 # 3

Methods


No methods listed for this paper. Add relevant methods here