Sparse Activation Maps for Interpreting 3D Object Detection

被引:7
作者
Chen, Qiuxiao [1 ]
Li, Pengfei [2 ]
Xu, Meng [1 ]
Qi, Xiaojun [1 ]
机构
[1] Utah State Univ, Logan, UT 84322 USA
[2] Univ Calif Riverside, Riverside, CA 92521 USA
来源
2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2021 | 2021年
关键词
D O I
10.1109/CVPRW53098.2021.00017
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We propose a technique to generate "visual explanations" for interpretability of volumetric-based 3D object detection networks. Specifically, we use the average pooling of weights to produce a Sparse Activation Map (SAM) which highlights the important regions of the 3D point cloud data. The SAMs is applicable to any volumetric-based models (model agnostic) to provide intuitive intermediate results at different layers to understand the complex network structures. The SAMs at the 3D feature map layer and the 2D feature map layer help to understand the effectiveness of neurons to capture the object information. The SAMs at the classification layer for each object class helps to understand the true positives and false positives of each network. The experimental results on the KITTI dataset demonstrate the visual observations of the SAM match the detection results of three volumetric-based models.
引用
收藏
页码:76 / 84
页数:9
相关论文
共 23 条
[1]  
Goldfeld Z., 2018, CoRR
[2]  
Hu P., 2020, P IEEE CVF C COMP VI, P11001
[3]   A survey of safety and trustworthiness of deep neural networks: Verification, testing, adversarial attack and defence, and interpretability? [J].
Huang, Xiaowei ;
Kroening, Daniel ;
Ruan, Wenjie ;
Sharp, James ;
Sun, Youcheng ;
Thamo, Emese ;
Wu, Min ;
Yi, Xinping .
COMPUTER SCIENCE REVIEW, 2020, 37
[4]  
Kim B, 2018, PR MACH LEARN RES, V80
[5]  
Koh PW, 2017, PR MACH LEARN RES, V70
[6]  
Mahendran A, 2015, PROC CVPR IEEE, P5188, DOI 10.1109/CVPR.2015.7299155
[7]   Extraction of rules from discrete-time recurrent neural networks [J].
Omlin, CW ;
Giles, CL .
NEURAL NETWORKS, 1996, 9 (01) :41-52
[8]  
Qi Charles R, 2017, ARXIV170602413
[9]  
Ribeiro M. T., 2018, AAAI C ART INT AAAI, DOI [DOI 10.1609/AAAI.V32I1.11491, 10.1609/aaai.v32i1.11491]
[10]   On the information bottleneck theory of deep learning [J].
Saxe, Andrew M. ;
Bansal, Yamini ;
Dapello, Joel ;
Advani, Madhu ;
Kolchinsky, Artemy ;
Tracey, Brendan D. ;
Cox, David D. .
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT, 2019, 2019 (12)