Decoder Generates Manufacturable Structures: A Framework for 3D-Printable Object Synthesis

TL;DR

This paper introduces a deep learning decoder framework that generates manufacturable 3D structures optimized for additive manufacturing, ensuring geometric validity and adherence to manufacturing constraints.

Contribution

It presents a novel neural decoder approach that learns to produce geometrically valid, printable 3D objects respecting manufacturing constraints, improving over naive generation methods.

Findings

Generated structures meet manufacturing constraints

Framework produces diverse, valid 3D geometries

Validated with practical 3D printing experiments

Abstract

This paper presents a novel decoder-based approach for generating manufacturable 3D structures optimized for additive manufacturing. We introduce a deep learning framework that decodes latent representations into geometrically valid, printable objects while respecting manufacturing constraints such as overhang angles, wall thickness, and structural integrity. The methodology demonstrates that neural decoders can learn complex mapping functions from abstract representations to valid 3D geometries, producing parts with significantly improved manufacturability compared to naive generation approaches. We validate the approach on diverse object categories and demonstrate practical 3D printing of decoder-generated structures.