3D Graphics News
The latest 3D Graphics coverage curated for game developers.
arXiv cs.GR details PartLLM, a multimodal model that treats 3D part segmentation as intent-conditioned generation. For graphics and tools teams, the practical draw is one system that can handle text-guided, …
A new motion-generation system aims to cover both humans and animals with one model, sidestepping the topology mismatch that usually forces separate pipelines. UniMo pairs a point-cloud representation with dynamic …
SceneHI brings high-resolution, 3D-consistent scene texturing with controllable illumination into a single generative pipeline. The system lifts priors from 2D diffusion models onto 3D objects, then bakes geometry-aware shadows directly …
MeshFlow generates triangle meshes directly with equivariant flow matching, sidestepping long autoregressive mesh sequences. The method respects face and vertex permutation symmetries and reports quality on par with leading mesh …
Hi-TOPS is a new 3D part decomposition method aimed at cleaner, editable mesh splits without semantic labels or 2D priors. It uses a hierarchical, topology-aware scoring prior to balance broad …
MeshFM turns 2D foundation-model features into 3D shape understanding with a feedforward pipeline that skips test-time optimization. Built from 2D supervision alone, it reaches performance comparable to 3D-supervised methods on …
A new 3D editing framework, CNS-Edit++, pushes neural shape manipulation beyond single categories and into foundation-model territory. It couples a global latent code with a 3D neural feature volume so …
A new mesh-generation method, LATO.2, splits the problem into two flows: first it generates vertices, then it derives connectivity from those realized points. That separation is aimed at fixing the …
MeshFlow replaces slow auto-regressive mesh tokenization with a continuous latent VAE plus a rectified-flow transformer, and it claims 18x faster generation than the fastest AR baseline. For graphics and tools …
CompoSE is a new 3D shape synthesis system that lets creators place coarse part layouts and get back editable, part-separated objects. It uses a diffusion transformer to balance local part …
The introduction of Feature-Aware Quadric Error Metric (FA-QEM) presents a significant advancement in mesh simplification, particularly beneficial for graphics programmers and artists. This method not only accelerates the simplification process …
The introduction of Denoising-GS offers a significant advancement in 3D Gaussian Splatting, particularly for graphics programmers focused on Novel View Synthesis. By addressing the noise introduced during optimization, this framework …
The introduction of Physics-Guided 3D Gaussian Splatting (PG-3DGS) marks a significant advancement in 3D scene generation. By integrating physics simulation with visual representation, developers can create objects that not only …
The exploration of 3D rotation representations is crucial for graphics programmers and developers in robotics and machine learning. This paper evaluates methods like quaternions and rotation matrices, highlighting their efficiency …
QuadBox introduces a significant leap in 3D Gaussian Splatting, enhancing rendering speeds by 1.85 times. This advancement is crucial for graphics programmers aiming for real-time performance in novel view synthesis, …
The Antipodal Method introduces a groundbreaking approach to computing 3D generalized winding numbers, achieving up to 22x speed improvements on CPUs. This advancement is crucial for graphics programmers working with …
A new system for indoor scene object placement leverages a Domain Specific Language (DSL) to improve placement predictions. This approach addresses limitations in existing data-driven methods, making it particularly relevant …
Kinematic Kitbashing introduces a novel optimization framework for synthesizing articulated 3D objects, which could significantly streamline workflows for graphics programmers and designers. By leveraging reusable parts and an abstract kinematic …
CasLayout introduces a novel approach to indoor scene synthesis, leveraging a cascaded diffusion framework to enhance spatial relationships and architectural integrity. This method significantly reduces data requirements while improving the …
The introduction of a structure-aware densification framework for 3D Gaussian Splatting significantly enhances convergence speed and reconstruction quality. This approach allows graphics programmers to achieve better texture representation without the …