MamTiff-CAD: Multi-Scale Latent Diffusion with Mamba+ for Complex Parametric Sequence

TL;DR

MamTiff-CAD introduces a multi-scale latent diffusion framework using Transformer and Mamba+ for generating complex, long parametric CAD command sequences, overcoming previous limitations in sequence length and model complexity.

Contribution

The paper presents a novel autoencoder with Mamba+ and Transformer, and a multi-scale diffusion model for long sequence CAD generation, advancing the state-of-the-art.

Findings

Achieves state-of-the-art performance in CAD sequence reconstruction.

Successfully generates long sequences up to 256 commands.

Demonstrates effectiveness on complex CAD models.

Abstract

Parametric Computer-Aided Design (CAD) is crucial in industrial applications, yet existing approaches often struggle to generate long sequence parametric commands due to complex CAD models' geometric and topological constraints. To address this challenge, we propose MamTiff-CAD, a novel CAD parametric command sequences generation framework that leverages a Transformer-based diffusion model for multi-scale latent representations. Specifically, we design a novel autoencoder that integrates Mamba+ and Transformer, to transfer parameterized CAD sequences into latent representations. The Mamba+ block incorporates a forget gate mechanism to effectively capture long-range dependencies. The non-autoregressive Transformer decoder reconstructs the latent representations. A diffusion model based on multi-scale Transformer is then trained on these latent embeddings to learn the distribution of long sequence commands. In addition, we also construct a dataset that consists of long parametric sequences, which is up to 256 commands for a single CAD model. Experiments demonstrate that MamTiff-CAD achieves state-of-the-art performance on both reconstruction and generation tasks, confirming its effectiveness for long sequence (60-256) CAD model generation.