Wavelet Latent Diffusion (Wala): Billion-Parameter 3D Generative Model with Compact Wavelet Encodings

• URL: http://arxiv.org/abs/2411.08017v1
• Date: Tue, 12 Nov 2024 18:49:06 GMT
• Title: Wavelet Latent Diffusion (Wala): Billion-Parameter 3D Generative Model with Compact Wavelet Encodings
• Authors: Aditya Sanghi, Aliasghar Khani, Pradyumna Reddy, Arianna Rampini, Derek Cheung, Kamal Rahimi Malekshan, Kanika Madan, Hooman Shayani,
• Abstract summary: Large-scale 3D generative models require substantial computational resources yet often fall short in capturing fine details and complex geometries at high resolutions. We introduce a novel approach called Wavelet Latent Diffusion, or WaLa, that encodes 3D shapes into compact latent encodings. Specifically, we compress a $2563$ signed distance field into a $123 times 4$ latent grid, achieving an impressive 2427x compression ratio with minimal loss of detail. Our models, both conditional and unconditional, contain approximately one billion parameters and successfully generate high-quality 3D shapes at $2563$
• Score: 15.2983201224858
• License:
• Abstract: Large-scale 3D generative models require substantial computational resources yet often fall short in capturing fine details and complex geometries at high resolutions. We attribute this limitation to the inefficiency of current representations, which lack the compactness required to model the generative models effectively. To address this, we introduce a novel approach called Wavelet Latent Diffusion, or WaLa, that encodes 3D shapes into wavelet-based, compact latent encodings. Specifically, we compress a $256^3$ signed distance field into a $12^3 \times 4$ latent grid, achieving an impressive 2427x compression ratio with minimal loss of detail. This high level of compression allows our method to efficiently train large-scale generative networks without increasing the inference time. Our models, both conditional and unconditional, contain approximately one billion parameters and successfully generate high-quality 3D shapes at $256^3$ resolution. Moreover, WaLa offers rapid inference, producing shapes within two to four seconds depending on the condition, despite the model's scale. We demonstrate state-of-the-art performance across multiple datasets, with significant improvements in generation quality, diversity, and computational efficiency. We open-source our code and, to the best of our knowledge, release the largest pretrained 3D generative models across different modalities.

Related papers

• Compress3D: a Compressed Latent Space for 3D Generation from a Single Image [27.53099431097921]
  Triplane autoencoder encodes 3D models into a compact triplane latent space to compress both the 3D geometry and texture information. We introduce a 3D-aware cross-attention mechanism, which utilizes low-resolution latent representations to query features from a high-resolution 3D feature volume. Our approach enables the generation of high-quality 3D assets in merely 7 seconds on a single A100 GPU.
  arXiv  Detail & Related papers  (2024-03-20T11:51:04Z)
• Pushing Auto-regressive Models for 3D Shape Generation at Capacity and Scalability [118.26563926533517]
  Auto-regressive models have achieved impressive results in 2D image generation by modeling joint distributions in grid space. We extend auto-regressive models to 3D domains, and seek a stronger ability of 3D shape generation by improving auto-regressive models at capacity and scalability simultaneously.
  arXiv  Detail & Related papers  (2024-02-19T15:33:09Z)
• LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creation [51.19871052619077]
  We introduce Large Multi-View Gaussian Model (LGM), a novel framework designed to generate high-resolution 3D models from text prompts or single-view images. We maintain the fast speed to generate 3D objects within 5 seconds while boosting the training resolution to 512, thereby achieving high-resolution 3D content generation.
  arXiv  Detail & Related papers  (2024-02-07T17:57:03Z)
• Make-A-Shape: a Ten-Million-scale 3D Shape Model [52.701745578415796]
  This paper introduces Make-A-Shape, a new 3D generative model designed for efficient training on a vast scale. We first innovate a wavelet-tree representation to compactly encode shapes by formulating the subband coefficient filtering scheme. We derive the subband adaptive training strategy to train our model to effectively learn to generate coarse and detail wavelet coefficients.
  arXiv  Detail & Related papers  (2024-01-20T00:21:58Z)
• GVP: Generative Volumetric Primitives [76.95231302205235]
  We present Generative Volumetric Primitives (GVP), the first pure 3D generative model that can sample and render 512-resolution images in real-time. GVP jointly models a number of primitives and their spatial information, both of which can be efficiently generated via a 2D convolutional network. Experiments on several datasets demonstrate superior efficiency and 3D consistency of GVP over the state-of-the-art.
  arXiv  Detail & Related papers  (2023-03-31T16:50:23Z)
• HoloDiffusion: Training a 3D Diffusion Model using 2D Images [71.1144397510333]
  We introduce a new diffusion setup that can be trained, end-to-end, with only posed 2D images for supervision. We show that our diffusion models are scalable, train robustly, and are competitive in terms of sample quality and fidelity to existing approaches for 3D generative modeling.
  arXiv  Detail & Related papers  (2023-03-29T07:35:56Z)
• Magic3D: High-Resolution Text-to-3D Content Creation [78.40092800817311]
  DreamFusion has recently demonstrated the utility of a pre-trained text-to-image diffusion model to optimize Neural Radiance Fields (NeRF) In this paper, we address these limitations by utilizing a two-stage optimization framework. Our method, dubbed Magic3D, can create high quality 3D mesh models in 40 minutes, which is 2x faster than DreamFusion.
  arXiv  Detail & Related papers  (2022-11-18T18:59:59Z)
• not-so-BigGAN: Generating High-Fidelity Images on Small Compute with Wavelet-based Super-Resolution [23.15896056344987]
  Nsb-GAN is a simple yet cost-effective two-step training framework for deep generative models. Wavelet-based down-sampling method preserves more structural information than pixel-based methods. On ImageNet 512x512, our model achieves a Fr'echet Inception Distance (FID) of 10.59 -- beating the baseline BigGAN model.
  arXiv  Detail & Related papers  (2020-09-09T17:29:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.