# Multi-Angle Point Cloud-VAE: Unsupervised Feature Learning for 3D Point Clouds from Multiple Angles by Joint Self-Reconstruction and Half-to-Half Prediction Source: [Multi-Angle Point Cloud-VAE: Unsupervised Feature Learning for 3D Point Clouds from Multiple Angles by Joint Self-Reconstruction and Half-to-Half Prediction]() Collection index: [llms.txt]() Data updated: 2026-01-04T03:09:13.036Z Authors: Zhizhong Han, Xiyang Wang, Yu-Shen Liu, Matthias Zwicker Published: 2019-07-30T02:17:12.000Z ## Abstract Unsupervised feature learning for point clouds has been vital for large-scale point cloud understanding. Recent deep learning based methods depend on learning global geometry from self-reconstruction. However, these methods are still suffering from ineffective learning of local geometry, which significantly limits the discriminability of learned features. To resolve this issue, we propose MAP-VAE to enable the learning of global and local geometry by jointly leveraging global and local self-supervision. To enable effective local self-supervision, we introduce multi-angle analysis for point clouds. In a multi-angle scenario, we first split a point cloud into a front half and a back half from each angle, and then, train MAP-VAE to learn to predict a back half sequence from the corresponding front half sequence. MAP-VAE performs this half-to-half prediction using RNN to simultaneously learn each local geometry and the spatial relationship among them. In addition, MAP-VAE also learns global geometry via self-reconstruction, where we employ a variational constraint to facilitate novel shape generation. The outperforming results in four shape analysis tasks show that MAP-VAE can learn more discriminative global or local features than the state-of-the-art methods. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Point Cloud Completion on ShapeNet Part Airplane](): Point Cloud Completion; ShapeNet Part Airplane; EMD (per point): 0.032328. - [Point Cloud Completion on ShapeNet Part Chair](): Point Cloud Completion; ShapeNet Part Chair; EMD (per point): 0.055696. - [Object Classification on ModelNet40 1.0 (test)](): Object Classification; ModelNet40 1.0 (test); Accuracy: 90.2. - [Object Classification on ModelNet10 (test)](): Object Classification; ModelNet10 (test); Accuracy: 94.8. - [3D Shape recognition on ModelNet10 (test)](): 3D shape recognition; ModelNet10 (test); Accuracy: 94.82. - [3D Object Part Segmentation on ShapeNet Part (test)](): 3D Object Part Segmentation; ShapeNet Part (test); mIoU: 67.95. - [3D Shape Classification on ModelNet40 (test)](): 3D Shape Classification; ModelNet40 (test); Accuracy: 90.15. - [Shape Classification on ModelNet40 (test)](): Shape classification; ModelNet40 (test); OA: not recorded. - [Object Classification on ModelNet40 (test)](): Object Classification; ModelNet40 (test); Accuracy: not recorded. ## Original sources and implementation [Original paper]() [PDF]()