Multi-view Convolutional Neural Networks for 3D Shape Recognition
This addresses the representation challenge in computer vision for 3D shape recognition, offering a more effective method that can also apply to hand-drawn sketches.
The paper tackles the problem of 3D shape recognition by comparing native 3D descriptors to view-based approaches, showing that a CNN trained on rendered 2D views achieves far higher accuracy than state-of-the-art 3D descriptors, with further improvements from multiple views and a novel multi-view CNN architecture.
A longstanding question in computer vision concerns the representation of 3D shapes for recognition: should 3D shapes be represented with descriptors operating on their native 3D formats, such as voxel grid or polygon mesh, or can they be effectively represented with view-based descriptors? We address this question in the context of learning to recognize 3D shapes from a collection of their rendered views on 2D images. We first present a standard CNN architecture trained to recognize the shapes' rendered views independently of each other, and show that a 3D shape can be recognized even from a single view at an accuracy far higher than using state-of-the-art 3D shape descriptors. Recognition rates further increase when multiple views of the shapes are provided. In addition, we present a novel CNN architecture that combines information from multiple views of a 3D shape into a single and compact shape descriptor offering even better recognition performance. The same architecture can be applied to accurately recognize human hand-drawn sketches of shapes. We conclude that a collection of 2D views can be highly informative for 3D shape recognition and is amenable to emerging CNN architectures and their derivatives.