REVIEW 6 cited by
LION: Latent Point Diffusion Models for 3D Shape Generation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Denoising diffusion models (DDMs) have shown promising results in 3D point cloud synthesis. To advance 3D DDMs and make them useful for digital artists, we require (i) high generation quality, (ii) flexibility for manipulation and applications such as conditional synthesis and shape interpolation, and (iii) the ability to output smooth surfaces or meshes. To this end, we introduce the hierarchical Latent Point Diffusion Model (LION) for 3D shape generation. LION is set up as a variational autoencoder (VAE) with a hierarchical latent space that combines a global shape latent representation with a point-structured latent space. For generation, we train two hierarchical DDMs in these latent spaces. The hierarchical VAE approach boosts performance compared to DDMs that operate on point clouds directly, while the point-structured latents are still ideally suited for DDM-based modeling. Experimentally, LION achieves state-of-the-art generation performance on multiple ShapeNet benchmarks. Furthermore, our VAE framework allows us to easily use LION for different relevant tasks: LION excels at multimodal shape denoising and voxel-conditioned synthesis, and it can be adapted for text- and image-driven 3D generation. We also demonstrate shape autoencoding and latent shape interpolation, and we augment LION with modern surface reconstruction techniques to generate smooth 3D meshes. We hope that LION provides a powerful tool for artists working with 3D shapes due to its high-quality generation, flexibility, and surface reconstruction. Project page and code: https://nv-tlabs.github.io/LION.
Forward citations
Cited by 6 Pith papers
-
Nexus: Native Mesh Generation with Diffusion
Nexus replaces autoregressive mesh serialization with two coupled diffusion models — octree vertex generation and a latent topology generator — claiming stronger geometry and perceptual quality on Objaverse and Toys4K.
-
BAG: Body-Aligned 3D Wearable Asset Generation
BAG generates body-aligned 3D wearable assets from a single image by conditioning multi-view diffusion on canonical body XYZ maps and refining alignment with Sim(3) optimization and physics simulation.
-
Coherent 3D Scene Diffusion From a Single RGB Image
A single RGB image is converted into a coherent 3D scene by denoising all object poses and shapes simultaneously with a diffusion model.
-
Reward Fine-Tuning Two-Step Diffusion Models via Learning Differentiable Latent-Space Surrogate Reward
A learned latent-space surrogate reward enables stable fine-tuning of one/two-step diffusion models with arbitrary, non-differentiable reward signals, outperforming policy-gradient baselines.
-
Wavelet Latent Diffusion (Wala): Billion-Parameter 3D Generative Model with Compact Wavelet Encodings
Wavelet Latent Diffusion (WaLa) shrinks 3D shapes to 6,912-variable latent codes and trains billion-parameter diffusion models that generate 256^3 geometry in 2-4 seconds, claiming state-of-the-art results.
-
Generative modeling assisted simulation of measurement-altered quantum criticality
A proposal to use a structure-preserving conditional diffusion model for simulating measurement-altered quantum criticality, backed only by locality data and not by a working generative model.
Discussion (0). Continue with ORCID to comment.