Corpus ID: 236469347

Surrogate Model-Based Explainability Methods for Point Cloud NNs

  title={Surrogate Model-Based Explainability Methods for Point Cloud NNs},
  author={Hanxiao Tan and Helena Kotthaus},
In the field of autonomous driving and robotics, point clouds are showing their excellent real-time performance as raw data from most of the mainstream 3D sensors. Therefore, point cloud neural networks have become a popular research direction in recent years. So far, however, there has been little discussion about the explainability of deep neural networks for point clouds. In this paper, we propose a point cloud-applicable explainability approach based on local surrogate model-based method to… Expand

Figures and Tables from this paper


3D Point Cloud Feature Explanations Using Gradient-Based Methods
Driven by the insight that 3D data is inherently sparse, this work visualises the features learnt by a voxel-based classification network and shows that these features are also sparse and can be pruned relatively easily, leading to more efficient neural networks. Expand
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
This paper designs a novel type of neural network that directly consumes point clouds, which well respects the permutation invariance of points in the input and provides a unified architecture for applications ranging from object classification, part segmentation, to scene semantic parsing. Expand
Frustum PointNets for 3D Object Detection from RGB-D Data
This work directly operates on raw point clouds by popping up RGBD scans and leverages both mature 2D object detectors and advanced 3D deep learning for object localization, achieving efficiency as well as high recall for even small objects. Expand
VoxNet: A 3D Convolutional Neural Network for real-time object recognition
  • Daniel Maturana, S. Scherer
  • Computer Science
  • 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
  • 2015
VoxNet is proposed, an architecture to tackle the problem of robust object recognition by integrating a volumetric Occupancy Grid representation with a supervised 3D Convolutional Neural Network (3D CNN). Expand
PointCloud Saliency Maps
A novel way of characterizing critical points and segments to build point-cloud saliency maps is proposed, and each saliency score can be efficiently measured by the corresponding gradient of the loss w.r.t the point under the spherical coordinates. Expand
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
A hierarchical neural network that applies PointNet recursively on a nested partitioning of the input point set and proposes novel set learning layers to adaptively combine features from multiple scales to learn deep point set features efficiently and robustly. Expand
PointHop: An Explainable Machine Learning Method for Point Cloud Classification
It is shown by experimental results that the PointHop method offers classification performance that is comparable with state-of-the-art methods while demanding much lower training complexity. Expand
Explaining Autonomous Driving by Learning End-to-End Visual Attention
This work proposes to train an imitation learning based agent equipped with an attention model that allows us to understand what part of the image has been deemed most important and leads to superior performance in a standard benchmark using the CARLA driving simulator. Expand
Visual Scene Understanding for Autonomous Driving Using Semantic Segmentation
This work suggests the use of a tiered approach, whose main component is a semantic segmentation model, over an end-to-end approach for an autonomous driving system. Expand
Detecting unexpected obstacles for self-driving cars: Fusing deep learning and geometric modeling
A principled Bayesian framework is presented to fuse the semantic and stereo-based detection results, and the mid-level Stixel representation is used to describe obstacles in a flexible, compact and robust manner. Expand