LegoFormer: Transformers for Block-by-Block Multi-view 3D Reconstruction

TL;DR

LegoFormer introduces a transformer-based approach for multi-view 3D reconstruction that shares information across views during all stages and predicts objects as structured sets, achieving competitive results and better interpretability.

Contribution

The paper presents LegoFormer, a novel transformer model that integrates view-sharing throughout the process and uses low-rank decomposition for structured object prediction.

Findings

Competitive performance on ShapeNet dataset

Enhanced interpretability via self-attention layers

Promising generalization to real-world data

Abstract

Most modern deep learning-based multi-view 3D reconstruction techniques use RNNs or fusion modules to combine information from multiple images after independently encoding them. These two separate steps have loose connections and do not allow easy information sharing among views. We propose LegoFormer, a transformer model for voxel-based 3D reconstruction that uses the attention layers to share information among views during all computational stages. Moreover, instead of predicting each voxel independently, we propose to parametrize the output with a series of low-rank decomposition factors. This reformulation allows the prediction of an object as a set of independent regular structures then aggregated to obtain the final reconstruction. Experiments conducted on ShapeNet demonstrate the competitive performance of our model with respect to the state of the art while having increased interpretability thanks to the self-attention layers. We also show promising generalization results to real data.