LegoFormer: Transformers for Block-by-Block Multi-view 3D Reconstruction
This work addresses multi-view 3D reconstruction for computer vision applications, presenting an incremental improvement by integrating transformers and low-rank decomposition.
The paper tackles the problem of multi-view 3D reconstruction by proposing LegoFormer, a transformer model that uses attention layers to share information among views and parametrizes outputs with low-rank decomposition factors, achieving competitive performance on ShapeNet with improved interpretability and promising generalization to real data.
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.