PointAugment: an Auto-Augmentation Framework for Point Cloud Classification
arXiv:2002.10876
Abstract
We present PointAugment, a new auto-augmentation framework that automatically optimizes and augments point cloud samples to enrich the data diversity when we train a classification network. Different from existing auto-augmentation methods for 2D images, PointAugment is sample-aware and takes an adversarial learning strategy to jointly optimize an augmentor network and a classifier network, such that the augmentor can learn to produce augmented samples that best fit the classifier. Moreover, we formulate a learnable point augmentation function with a shape-wise transformation and a point-wise displacement, and carefully design loss functions to adopt the augmented samples based on the learning progress of the classifier. Extensive experiments also confirm PointAugment's effectiveness and robustness to improve the performance of various networks on shape classification and retrieval.
arXiv commentCamera-Ready Version for CVPR 2020 (Oral); code is https://github.com/liruihui/PointAugment/
References in corpus (14)
- mixup: Beyond Empirical Risk Minimization
- NIPS 2016 Tutorial: Generative Adversarial Networks
- Smart Augmentation - Learning an Optimal Data Augmentation Strategy
- Fractional Max-Pooling
- Learning to Compose Domain-Specific Transformations for Data Augmentation
- Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules
- Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities
- A Bayesian Data Augmentation Approach for Learning Deep Models
- Snapshot Ensembles: Train 1, get M for free
- Adversarial AutoAugment
- PU-Net: Point Cloud Upsampling Network
- LassoNet: Deep Lasso-Selection of 3D Point Clouds
- Dataset Augmentation in Feature Space
- PointCleanNet: Learning to Denoise and Remove Outliers from Dense Point Clouds