PASTA: Controllable Part-Aware Shape Generation with Autoregressive Transformers

• URL: http://arxiv.org/abs/2407.13677v1
• Date: Thu, 18 Jul 2024 16:52:45 GMT
• Title: PASTA: Controllable Part-Aware Shape Generation with Autoregressive Transformers
• Authors: Songlin Li, Despoina Paschalidou, Leonidas Guibas,
• Abstract summary: PASTA is an autoregressive transformer architecture for generating high quality 3D shapes. Our model generates 3D shapes that are both more realistic and diverse than existing part-based and non part-based methods.
• Score: 5.7181794813117754
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The increased demand for tools that automate the 3D content creation process led to tremendous progress in deep generative models that can generate diverse 3D objects of high fidelity. In this paper, we present PASTA, an autoregressive transformer architecture for generating high quality 3D shapes. PASTA comprises two main components: An autoregressive transformer that generates objects as a sequence of cuboidal primitives and a blending network, implemented with a transformer decoder that composes the sequences of cuboids and synthesizes high quality meshes for each object. Our model is trained in two stages: First we train our autoregressive generative model using only annotated cuboidal parts as supervision and next, we train our blending network using explicit 3D supervision, in the form of watertight meshes. Evaluations on various ShapeNet objects showcase the ability of our model to perform shape generation from diverse inputs \eg from scratch, from a partial object, from text and images, as well size-guided generation, by explicitly conditioning on a bounding box that defines the object's boundaries. Moreover, as our model considers the underlying part-based structure of a 3D object, we are able to select a specific part and produce shapes with meaningful variations of this part. As evidenced by our experiments, our model generates 3D shapes that are both more realistic and diverse than existing part-based and non part-based methods, while at the same time is simpler to implement and train.

Related papers

• MeshAnything: Artist-Created Mesh Generation with Autoregressive Transformers [76.70891862458384]
  We introduce MeshAnything, a model that treats mesh extraction as a generation problem. By converting 3D assets in any 3D representation into AMs, MeshAnything can be integrated with various 3D asset production methods. Our method generates AMs with hundreds of times fewer faces, significantly improving storage, rendering, and simulation efficiencies.
  arXiv  Detail & Related papers  (2024-06-14T16:30:25Z)
• CLAY: A Controllable Large-scale Generative Model for Creating High-quality 3D Assets [43.315487682462845]
  CLAY is a 3D geometry and material generator designed to transform human imagination into intricate 3D digital structures. At its core is a large-scale generative model composed of a multi-resolution Variational Autoencoder (VAE) and a minimalistic latent Diffusion Transformer (DiT) We demonstrate using CLAY for a range of controllable 3D asset creations, from sketchy conceptual designs to production ready assets with intricate details.
  arXiv  Detail & Related papers  (2024-05-30T05:57:36Z)
• NeuSDFusion: A Spatial-Aware Generative Model for 3D Shape Completion, Reconstruction, and Generation [52.772319840580074]
  3D shape generation aims to produce innovative 3D content adhering to specific conditions and constraints. Existing methods often decompose 3D shapes into a sequence of localized components, treating each element in isolation. We introduce a novel spatial-aware 3D shape generation framework that leverages 2D plane representations for enhanced 3D shape modeling.
  arXiv  Detail & Related papers  (2024-03-27T04:09:34Z)
• Object-Centric Domain Randomization for 3D Shape Reconstruction in the Wild [22.82439286651921]
  One of the biggest challenges in single-view 3D shape reconstruction in the wild is the scarcity of 3D shape, 2D image>-paired data from real-world environments. Inspired by remarkable achievements via domain randomization, we propose ObjectDR which synthesizes such paired data via a random simulation of visual variations in object appearances and backgrounds.
  arXiv  Detail & Related papers  (2024-03-21T16:40:10Z)
• ComboVerse: Compositional 3D Assets Creation Using Spatially-Aware Diffusion Guidance [76.7746870349809]
  We present ComboVerse, a 3D generation framework that produces high-quality 3D assets with complex compositions by learning to combine multiple models. Our proposed framework emphasizes spatial alignment of objects, compared with standard score distillation sampling.
  arXiv  Detail & Related papers  (2024-03-19T03:39:43Z)
• 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)
• Pushing the Limits of 3D Shape Generation at Scale [65.24420181727615]
  We present a significant breakthrough in 3D shape generation by scaling it to unprecedented dimensions. We have developed a model with an astounding 3.6 billion trainable parameters, establishing it as the largest 3D shape generation model to date, named Argus-3D.
  arXiv  Detail & Related papers  (2023-06-20T13:01:19Z)
• Learning to Generate 3D Shapes from a Single Example [28.707149807472685]
  We present a multi-scale GAN-based model designed to capture the input shape's geometric features across a range of spatial scales. We train our generative model on a voxel pyramid of the reference shape, without the need of any external supervision or manual annotation. The resulting shapes present variations across different scales, and at the same time retain the global structure of the reference shape.
  arXiv  Detail & Related papers  (2022-08-05T01:05:32Z)
• Towards Realistic 3D Embedding via View Alignment [53.89445873577063]
  This paper presents an innovative View Alignment GAN (VA-GAN) that composes new images by embedding 3D models into 2D background images realistically and automatically. VA-GAN consists of a texture generator and a differential discriminator that are inter-connected and end-to-end trainable.
  arXiv  Detail & Related papers  (2020-07-14T14:45:00Z)

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.