Scaling Mesh Generation via Compressive Tokenization
This addresses the challenge of scaling mesh generation for applications requiring intricate details, though it appears incremental as it builds on existing tokenization methods.
The paper tackles the problem of generating high-detail meshes by proposing Blocked and Patchified Tokenization (BPT), a compressive mesh representation that reduces sequence length by about 75% and enables generation of meshes with over 8k faces, achieving state-of-the-art performance suitable for direct product use.
We propose a compressive yet effective mesh representation, Blocked and Patchified Tokenization (BPT), facilitating the generation of meshes exceeding 8k faces. BPT compresses mesh sequences by employing block-wise indexing and patch aggregation, reducing their length by approximately 75\% compared to the original sequences. This compression milestone unlocks the potential to utilize mesh data with significantly more faces, thereby enhancing detail richness and improving generation robustness. Empowered with the BPT, we have built a foundation mesh generative model training on scaled mesh data to support flexible control for point clouds and images. Our model demonstrates the capability to generate meshes with intricate details and accurate topology, achieving SoTA performance on mesh generation and reaching the level for direct product usage.