Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptions (2022-01-28T00:00:00.000000Z)

TL;DR

This paper presents ModelNet40-C, the first comprehensive benchmark on 3D point cloud corruption robustness, consisting of 15 common and realistic corruptions, and unveils that Transformer-based architectures with proper training recipes achieve the strongest robustness.

Abstract

Deep neural networks on 3D point cloud data have been widely used in the real world, especially in safety-critical applications. However, their robustness against corruptions is less studied. In this paper, we present ModelNet40-C, the first comprehensive benchmark on 3D point cloud corruption robustness, consisting of 15 common and realistic corruptions. Our evaluation shows a significant gap between the performances on ModelNet40 and ModelNet40-C for state-of-the-art (SOTA) models. To reduce the gap, we propose a simple but effective method by combining PointCutMix-R and TENT after evaluating a wide range of augmentation and test-time adaptation strategies. We identify a number of critical insights for future studies on corruption robustness in point cloud recognition. For instance, we unveil that Transformer-based architectures with proper training recipes achieve the strongest robustness. We hope our in-depth analysis will motivate the development of robust training strategies or architecture designs in the 3D point cloud domain. Our codebase and dataset are included in https://github.com/jiachens/ModelNet40-C

Authors

B. Kailkhura

4 papers

Chaowei Xiao

4 papers

Zhiding Yu

10 papers

TL;DR

Abstract

Authors

References71 items

3D Medical Point Transformer: Introducing Convolution to Attention Networks for Medical Point Cloud Analysis

Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselines

Are Transformers More Robust Than CNNs?

Interpreting Representation Quality of DNNs for 3D Point Cloud Processing

Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline

Verifiability and Predictability: Interpreting Utilities of Network Architectures for Point Cloud Processing

3D Shape Generation and Completion through Point-Voxel Diffusion

Defending Against Image Corruptions Through Adversarial Augmentations

PointGuard: Provably Robust 3D Point Cloud Classification

On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness

Regularization Strategy for Point Cloud via Rigidly Mixed Sample

PointCutMix: Regularization Strategy for Point Cloud Classification

PCT: Point cloud transformer

WILDS: A Benchmark of in-the-Wild Distribution Shifts

RobustBench: a standardized adversarial robustness benchmark

PointMixup: Augmentation for Point Clouds

Improving robustness against common corruptions by covariate shift adaptation

Towards Robust LiDAR-based Perception in Autonomous Driving: General Black-box Adversarial Sensor Attack and Countermeasures

Center-based 3D Object Detection and Tracking

Self-Robust 3D Point Recognition via Gather-Vector Guidance

A morphing-Based 3D point cloud reconstruction framework for medical image processing

Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks

On Isometry Robustness of Deep 3D Point Cloud Models Under Adversarial Attacks

PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection

AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty

PyTorch: An Imperative Style, High-Performance Deep Learning Library

Revisiting Point Cloud Classification: A New Benchmark Dataset and Classification Model on Real-World Data

3D Point Cloud Compression: A Survey

Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming

3D point cloud analysis for automatic inspection of aeronautical mechanical assemblies

Deep Closest Point: Learning Representations for Point Cloud Registration

KPConv: Flexible and Deformable Convolution for Point Clouds

Relation-Shape Convolutional Neural Network for Point Cloud Analysis

Benchmarking Neural Network Robustness to Common Corruptions and Perturbations

On Evaluating Adversarial Robustness

Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud Classifiers

DUP-Net: Denoiser and Upsampler Network for 3D Adversarial Point Clouds Defense

PointRCNN: 3D Object Proposal Generation and Detection From Point Cloud

PointConv: Deep Convolutional Networks on 3D Point Clouds

Generating 3D Adversarial Point Clouds

TOF Lidar Development in Autonomous Vehicle

AutoAugment: Learning Augmentation Policies from Data

Open3D: A Modern Library for 3D Data Processing

Dynamic Graph CNN for Learning on Point Clouds

PointCNN: Convolution On X-Transformed Points

Mining Point Cloud Local Structures by Kernel Correlation and Graph Pooling

Large-Scale Point Cloud Semantic Segmentation with Superpoint Graphs

Improved Regularization of Convolutional Neural Networks with Cutout

PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space

ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes

Explorations for real-time point cloud rendering of natural scenes in virtual reality

PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation

Point Cloud Noise and Outlier Removal for Image-Based 3D Reconstruction

Deep Residual Learning for Image Recognition

ShapeNet: An Information-Rich 3D Model Repository

VoxNet: A 3D Convolutional Neural Network for real-time object recognition

Voting for Voting in Online Point Cloud Object Detection

A Review of Point Cloud Registration Algorithms for Mobile Robotics

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Adam: A Method for Stochastic Optimization

3D ShapeNets: A deep representation for volumetric shapes

Efficient geometrical parametrisation techniques of interfaces for reduced-order modelling: application to fluid–structure interaction coupling problems

3D point cloud segmentation: A survey

Are we ready for autonomous driving? The KITTI vision benchmark suite

Free-form deformation of solid geometric models

Adversarially Robust 3D Point Cloud Recognition Using Self-Supervisions

Tent: Fully Test-Time Adaptation by Entropy Minimization

On Adversarial Robustness of 3D Point Cloud Classification under Adaptive Attacks

Lidar with Velocity: Motion Distortion Correction of Point Clouds from Oscillating Scanning Lidars

Anomalous Example Detection in Deep Learning: A Survey

The Shapley Value

Field of Study

Journal Information

Name

Volume

Venue Information