Transformation-Equivariant Network Fused With Multi-Stage Cascade Attention for Point Cloud Object Detection

Abstract

Object detection of unmanned firefighting vehicles faces challenges such as strong electromagnetic interference, drastic lighting changes and dynamic object variations. To address these issues, we propose a two-stage 3D point cloud object detection algorithm called TED-CasA-Fusion. The first stage uses the transformation-equivariant detector backbone that explicitly models rotation/reflection equivariance via weight-sharing sparse convolutions, which improves detection robustness to dynamically transformed objects. The second stage introduces a cascade attention-based multistage refinement network that aggregates cross-stage object features through cascade attention modules, which effectively enhances feature representation for multiscale objects. Furthermore, the second stage integrates weighted bounding box voting to address training imbalance due to dense nearby and sparse distant point distributions, thereby improving detection accuracy for distant and sparse targets. Comparative experiments were conducted on the KITTI dataset and a self-collected firefighting dataset between the proposed algorithm and some state-of-the-art algorithms. Results show that the proposed algorithm achieves the best 3D detection accuracy for hard-category objects on the KITTI dataset and also outperforms other detection approaches on the firefighting dataset. This work offers an efficient and reliable solution to environmental perception of unmanned firefighting vehicles.

Keywords

3D object detection / cascade attention / substation / transformation equivariance / unmanned firefighting vehicle

Acknowledgements

This work was supported by the National Key R&D Program of China (Grant 2022YFD2000400).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data are available from the authors upon request.

References

[1] F. H. Panahi, F. H. Panahi, and T. Ohtsuki, “A Reinforcement Learning-Based Fire Warning and Suppression System Using Unmanned Aerial Vehicles,” IEEE Transactions on Instrumentation and Measurement 72 (2022): 5500216, https://doi.org/10.1109/TIM.2022.3227558.

[2] L. Chen, Y. Xie, Y. He, et al., Autonomous Mining Through Cooperative Driving and Operations Enabled by Parallel Intelligence,” Communications Engineering 3, no. 1 (2024): 75, https://doi.org/10.1038/s44172-024-00220-5.

[3] S. Wang, Y. Zhang, T. H. Hsieh, W. Liu, F. Yin, and B. Liu, “Fire Situation Detection Method for Unmanned Fire-Fighting Vessel Based on Coordinate Attention Structure-Based Deep Learning Network,” Ocean Engineering 266 (2022): 113208, https://doi.org/10.1016/j.oceaneng.2022.113208.

[4] D. Ren, Z. Wang, H. Sun, L. Liu, W. Wang, and J. Zhang, “Salience Feature Guided Decoupling Network for UAV Forests Flame Detection,” Expert Systems with Applications 270 (2025): 126414, https://doi.org/10.1016/j.eswa.2025.126414.

[5] S. P. H. Boroujeni and A. Razi, “IC-GAN: An Improved Conditional Generative Adversarial Network for RGB-to-IR Image Translation With Applications to Forest Fire Monitoring,” Expert Systems with Applications 238 (2024): 121962, https://doi.org/10.1016/j.eswa.2023.121962.

[6] F. Schenkel, T. Schultze, C. Baer, et al., Radar-Based Smoke Detection at Millimeter Wave Frequencies: An Experimental Study,” in Proceedings of the IEEE/MTT-S International Microwave Symposium-IMS (2024), 887-890, https://doi.org/10.1109/IMS40175.2024.10600292.

[7] A. Bystrov, L. Daniel, E. Hoare, F. Norouzian, M. Cherniakov, and M. Gashinova, “Experimental Evaluation of 79 and 300 GHz Radar Performance in Fire Environments,” Sensors 21, no. 2 (2021): 439, https://doi.org/10.3390/s21020439.

[8] Q. Chen, Y. Xie, S. Guo, J. Bai, and Q. Shu, “Sensing System of Environmental Perception Technologies for Driverless Vehicle: A Review of State of the Art and Challenges,” Sensors and Actuators A: Physical 319 (2021): 112566, https://doi.org/10.1016/j.sna.2021.112566.

[9] S. Wang, M. N. Nikolić, T. L. Lam, Q. Gao, R. Ding, and T. Zhang, “Robot Manipulation Based on Embodied Visual Perception: A Survey,” CAAI Transactions on Intelligence Technology 10, no. 4 (2025): 945-958, https://doi.org/10.1049/cit2.70022.

[10] S. Yao, R. Guan, X. Huang, et al., Radar-Camera Fusion for Object Detection and Semantic Segmentation in Autonomous Driving: A Comprehensive Review,” IEEE Transactions on Intelligent Vehicles 9, no. 1 (2023): 2094-2128, https://doi.org/10.1109/TIV.2023.3307157.

[11] G. Zamanakos, L. Tsochatzidis, A. Amanatiadis, and I. Pratikakis, “A Comprehensive Survey of LIDAR-Based 3D Object Detection Methods With Deep Learning for Autonomous Driving,” Computers & Graphics 99 (2021): 153-181, https://doi.org/10.1016/j.cag.2021.07.003.

[12] Z. Yang, Y. Sun, S. Liu, et al., 3DSSD: Point-Based 3D Single Stage Object Detector,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2020), 11040-11048.

[13] W. Zheng, W. Tang, L. Jiang, et al., SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2021), 14494-14503.

[14] A. Liang, H. Hua, J. Fang, et al., PDM-SSD: Single-Stage Three-Dimensional Object Detector With Point Dilation,” preprint, arXiv: 2502.07822 (2025).

[15] Y. Yan, Y. Mao, and B. Li, “Second: Sparsely Embedded Convolutional Detection,” Sensors 18, no. 10 (2018): 3337, https://doi.org/10.3390/s18103337.

[16] A. H. Lang, S. Vora, H. Caesar, et al., PointPillars: Fast Encoders for Object Detection From Point Clouds,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2019), 12697-12705.

[17] Y. Chen, J. Liu, X. Zhang, et al., VoxelNeXt: Fully Sparse VoxelNet for 3D Object Detection and Tracking,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023), 21674-21683.

[18] R. Ma, C. Chen, B. Yang, et al., CG-SSD: Corner Guided Single Stage 3D Object Detection From LiDAR Point Cloud,” ISPRS Journal of Photogrammetry and Remote Sensing 191 (2022): 33-48, https://doi.org/10.1016/j.isprsjprs.2022.07.006.

[19] Z. Leng, P. Sun, T. He, et al., PVTransformer: Point-to-Voxel Transformer for Scalable 3D Object Detection,” in Proceedings of the IEEE International Conference on Robotics and Automation (2024), 4238-4244.

[20] J. Li, S. Dong, L. Ding, and T. Xu, “MsSVT++: Mixed-Scale Sparse Voxel Transformer With Center Voting for 3D Object Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence 46, no. 5 (2023): 3736-3752, https://doi.org/10.1109/TPAMI.2023.3345880.

[21] J. Lu, G. Ji, C. Cai, and K. Qin, “A Transformer Based on Voxel Spatial-Channel Attention for 3D Object Detection,” Pattern Recognition Letters 195 (2025): 37-43, https://doi.org/10.1016/j.patrec.2025.04.034.

[22] B. Wang, C. Xia, X. Gao, et al., Instance-Aware Sampling and Voxel-Transformer Encoding for Single-Stage 3D Object Detection,” Digital Signal Processing 162 (2025): 105171, https://doi.org/10.1016/j.dsp.2025.105171.

[23] H. Liu, J. Du, Y. Zhang, H. Zhang, and J. Zeng, “Syn-Aug: An Effective and General Synchronous Data Augmentation Framework for 3D Object Detection,” CAAI Transactions on Intelligence Technology 10, no. 3 (2025): 912-928, https://doi.org/10.1049/cit2.70001.

[24] S. Shi, X. Wang, and H. Li, “PointRCNN: 3D Object Proposal Generation and Detection From Point Cloud,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2019), 770-779.

[25] J. Deng, S. Shi, P. Li, W. Zhou, and Y. Zhang, “Voxel R-CNN: Towards High Performance Voxel-Based 3D Object Detection,” Proceedings of the AAAI Conference on Artificial Intelligence 35, no. 2 (2021): 1201-1209, https://doi.org/10.1609/aaai.v35i2.16207.

[26] H. Wang, L. Tao, Y. Peng, Z. Chen, and Y. Zhang, “Voxel RCNN-HA: A Point Cloud Multiobject Detection Algorithm With Hybrid Anchors for Autonomous Driving,” IEEE Transactions on Transportation Electrification 10, no. 3 (2023): 7286-7296, https://doi.org/10.1109/TTE.2023.3346375.

[27] F. Sun, G. Tong, and Y. Song, “Efficient Flexible Voxel-Based Two-Stage Network for 3D Object Detection in Autonomous Driving,” Applied Soft Computing 162 (2024): 111856, https://doi.org/10.1016/j.asoc.2024.111856.

[28] S. Shi, C. Guo, L. Jiang, et al., PV-RCNN: Point-Voxel Feature Set Abstraction for 3D Object Detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2020), 10529-10538.

[29] S. Shi, L. Jiang, J. Deng, et al., PV-RCNN++: Point-Voxel Feature Set Abstraction With Local Vector Representation for 3D Object Detection,” International Journal of Computer Vision 131, no. 2 (2023): 531-551, https://doi.org/10.1007/s11263-022-01710-9.

[30] S. Shi, Z. Wang, J. Shi, X. Wang, and H. Li, “From Points to Parts: 3D Object Detection From Point Cloud With Part-Aware and Part-Aggregation Network,” IEEE Transactions on Pattern Analysis and Machine Intelligence 43, no. 8 (2020): 2647-2664, https://doi.org/10.1109/TPAMI.2020.2977026.

[31] D. Shanmugam, D. Blalock, G. Balakrishnan, et al., Better Aggregation in Test-Time Augmentation,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (2021), 1214-1223.

[32] H. X. Yu, J. Wu, and L. Yi, “Rotationally Equivariant 3D Object Detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2022), 1456-1464.

[33] T. S. Cohen, M. Geiger, J. Köhler, et al., Spherical CNNs,” in International Conference on Learning Representations (2018).

[34] M. Weiler and G. Cesa, “General E(2)-Equivariant Steerable CNNs,” in Proceedings of the 33rd International Conference on Neural Information Processing Systems (2019), 14357-14368.

[35] M. Weiler, M. Geiger, M. Welling, et al., 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data,” in Proceedings of the 32nd International Conference on Neural Information Processing Systems (2018), 10402-10413.

[36] H. Chen, S. Liu, W. Chen, et al., Equivariant Point Network for 3D Point Cloud Analysis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2021), 14514-14523.

[37] J. Lin, H. Li, K. Chen, et al., Sparse Steerable Convolutions: An Efficient Learning of SE(3)-Equivariant Features for Estimation and Tracking of Object Poses in 3D Space,” in Proceedings of the 35th International Conference on Neural Information Processing Systems (2021), 16779-16790.

[38] J. Han, J. Ding, N. Xue, et al., ReDet: A Rotation-Equivariant Detector for Aerial Object Detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2021), 2786-2795.

[39] H. Wu, J. Deng, C. Wen, X. Li, C. Wang, and J. Li, “CasA: A Cascade Attention Network for 3-D Object Detection From LiDAR Point Clouds,” IEEE Transactions on Geoscience and Remote Sensing 60 (2022): 1-11, https://doi.org/10.1109/TGRS.2022.3203163.

[40] Z. Liu, J. Du, F. Tian, and J. Wen, “MR-CNN: A Multi-Scale Region-Based Convolutional Neural Network for Small Traffic Sign Recognition,” IEEE Access 7 (2019): 57120-57128, https://doi.org/10.1109/ACCESS.2019.2913882.

[41] Z. Cai and N. Vasconcelos, “Cascade R-CNN: Delving Into High Quality Object Detection,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2018), 6154-6162.

[42] Z. Cai and N. Vasconcelos, “Cascade R-CNN: High Quality Object Detection and Instance Segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence 43, no. 5 (2019): 1483-1498, https://doi.org/10.1109/TPAMI.2019.2956516.

[43] S. Gidaris and N. Komodakis, “Object Detection via a Multi-Region and Semantic Segmentation-Aware CNN Model,” in Proceedings of the IEEE International Conference on Computer Vision (2015), 1134-1142.

[44] S. Zagoruyko, A. Lerer, T. Y. Lin, et al., A MultiPath Network for Object Detection,” preprint, arXiv:1604.02135 (2016).

[45] Y. Tian, X. Zhang, X. Wang, et al., ACF-Net: Asymmetric Cascade Fusion for 3D Detection With LiDAR Point Clouds and Images,” IEEE Transactions on Intelligent Vehicles 9, no. 2 (2023): 3360-3371, https://doi.org/10.1109/TIV.2023.3341223.

[46] Q. Meng, W. Wang, T. Zhou, J. Shen, Y. Jia, and L. Van Gool, “Towards a Weakly Supervised Framework for 3D Point Cloud Object Detection and Annotation,” IEEE Transactions on Pattern Analysis and Machine Intelligence 44, no. 8 (2021): 4454-4468, https://doi.org/10.1109/TPAMI.2021.3063611.

[47] H. Wu, C. Wen, W. Li, X. Li, R. Yang, and C. Wang, “Transformation-Equivariant 3D Object Detection for Autonomous Driving,” Proceedings of the AAAI Conference on Artificial Intelligence 37, no. 3 (2023): 2795-2802, https://doi.org/10.1609/aaai.v37i3.25380.

[48] B. Graham, M. Engelcke, and L. Van Der Maaten, “3D Semantic Segmentation With Submanifold Sparse Convolutional Networks,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2018), 9224-9232.

[49] M. Weiler, F. A. Hamprecht, and M. Storath, “Learning Steerable Filters for Rotation Equivariant CNNs,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2018), 849-858.

[50] D. Laptev, N. Savinov, J. M. Buhmann, et al., TI-Pooling: Transformation-Invariant Pooling for Feature Learning in Convolutional Neural Networks,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2016), 289-297.

[51] A. Vaswani, N. Shazeer, N. Parmar, et al., Attention Is all You Need,” in Proceedings of the 31th International Conference on Neural Information Processing Systems (2017), 30.

[52] C. He, H. Zeng, J. Huang, et al., Structure Aware Single-Stage 3D Object Detection From Point Cloud,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2020), 11873-11882.

[53] A. Geiger, P. Lenz, and R. Urtasun, “Are We Ready for Autonomous Driving? The Kitti Vision Benchmark Suite,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (2012), 3354-3361.

[54] E. Li, S. Wang, C. Li, et al., SUSTech POINTS: A Portable 3D Point Cloud Interactive Annotation Platform System,” in Proceedings of the IEEE Intelligent Vehicles Symposium (2020), 1108-1115.

[55] G. Shi, R. Li, and C. Ma, “PillarNet: Real-Time and High-Performance Pillar-Based 3D Object Detection,” in European Conference on Computer Vision (2022), 35-52.

[56] T. Yin, X. Zhou, and P. Krahenbuhl, “Center-Based 3D Object Detection and Tracking,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2021), 11784-11793.