Search papers, labs, and topics across Lattice.
The paper introduces Easy3E, a feed-forward framework for editing 3D assets from a single view using a TRELLIS generative backbone. It addresses 3D consistency by introducing Voxel FlowEdit, an edit-driven flow in the sparse voxel latent space for globally consistent deformation, and enhances appearance fidelity with a normal-guided single-to-multi-view generation module. The method achieves fast, globally consistent, and high-fidelity 3D model editing without iterative optimization.
Forget iterative optimization: now you can edit 3D models in a single feedforward pass with globally consistent deformations and high-fidelity textures.
Existing 3D editing methods rely on computationally intensive scene-by-scene iterative optimization and suffer from multi-view inconsistency. We propose an effective and fully feedforward 3D editing framework based on the TRELLIS generative backbone, capable of modifying 3D models from a single editing view. Our framework addresses two key issues: adapting training-free 2D editing to structured 3D representations, and overcoming the bottleneck of appearance fidelity in compressed 3D features. To ensure geometric consistency, we introduce Voxel FlowEdit, an edit-driven flow in the sparse voxel latent space that achieves globally consistent 3D deformation in a single pass. To restore high-fidelity details, we develop a normal-guided single to multi-view generation module as an external appearance prior, successfully recovering high-frequency textures. Experiments demonstrate that our method enables fast, globally consistent, and high-fidelity 3D model editing.
