# Iterative Transformer Network for 3D Point Cloud

arXiv: https://arxiv.org/abs/1811.11209 · HTML: https://mlsanity.com/paper/1811.11209/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Nov 2018.

## Built on by (most cited first, 11 of 11)

- [Learning general and distinctive 3D local deep descriptors for point cloud registration](https://mlsanity.com/paper/2105.10382/) (2021)
- [R-PointHop: A Green, Accurate, and Unsupervised Point Cloud Registration Method](https://mlsanity.com/paper/2103.08129/) (2021)
- [Self-Supervised Point Cloud Completion via Inpainting](https://mlsanity.com/paper/2111.10701/) (2021)
- [EPiC: Ensemble of Partial Point Clouds for Robust Classification](https://mlsanity.com/paper/2303.11419/) (2023)
- [Adjoint Rigid Transform Network: Task-conditioned Alignment of 3D Shapes](https://mlsanity.com/paper/2102.01161/) (2021)
- [Generative Models for 3D Point Clouds](https://mlsanity.com/paper/2302.13408/) (2023)
- [The Devils in the Point Clouds: Studying the Robustness of Point Cloud Convolutions](https://mlsanity.com/paper/2101.07832/) (2021)
- [PointTree: Transformation-Robust Point Cloud Encoder with Relaxed K-D Trees](https://mlsanity.com/paper/2208.05962/) (2022)
- [Dual-Neighborhood Deep Fusion Network for Point Cloud Analysis](https://mlsanity.com/paper/2108.09228/) (2021)
- [Point-Selection Fine-Tuning Framework for Robust Point Cloud Classification](https://mlsanity.com/paper/2607.19711/) (2026)
- [IRGNN: Efficient Invariant Radar Graph Neural Network for Radar Point Cloud Object Detection](https://mlsanity.com/paper/2608.14394/) (2026)

## Abstract (arXiv)

3D point cloud is an efficient and flexible representation of 3D structures. Recently, neural networks operating on point clouds have shown superior performance on 3D understanding tasks such as shape classification and part segmentation. However, performance on such tasks is evaluated on complete shapes aligned in a canonical frame, while real world 3D data are partial and unaligned. A key challenge in learning from partial, unaligned point cloud data is to learn features that are invariant or equivariant with respect to geometric transformations. To address this challenge, we propose the Iterative Transformer Network (IT-Net), a network module that canonicalizes the pose of a partial object with a series of 3D rigid transformations predicted in an iterative fashion. We demonstrate the efficacy of IT-Net as an anytime pose estimator from partial point clouds without using complete object models. Further, we show that IT-Net achieves superior performance over alternative 3D transformer networks on various tasks, such as partial shape classification and object part segmentation.