Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
153 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Hierarchical Insights: Exploiting Structural Similarities for Reliable 3D Semantic Segmentation (2404.06124v3)

Published 9 Apr 2024 in cs.CV, cs.AI, and cs.RO

Abstract: Safety-critical applications such as autonomous driving require robust 3D environment perception algorithms capable of handling diverse and ambiguous surroundings. The predictive performance of classification models is heavily influenced by the dataset and the prior knowledge provided by the annotated labels. While labels guide the learning process, they often fail to capture the inherent relationships between classes that are naturally understood by humans. We propose a training strategy for a 3D LiDAR semantic segmentation model that learns structural relationships between classes through abstraction. This is achieved by implicitly modeling these relationships using a learning rule for hierarchical multi-label classification (HMC). Our detailed analysis demonstrates that this training strategy not only improves the model's confidence calibration but also retains additional information useful for downstream tasks such as fusion, prediction, and planning.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (41)
  1. Wikipedia, “Death of Elaine Herzberg — Wikipedia, the free encyclopedia,” https://en.wikipedia.org/wiki/Death˙of˙Elaine˙Herzberg, 2023, [Online; accessed 27-July-2023].
  2. J. Behley et al., “SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR Sequences,” in IEEE ICCV, 2019.
  3. A. Milioto et al., “RangeNet ++: Fast and Accurate LiDAR Semantic Segmentation,” in IEEE IROS, 2019.
  4. T. Cortinhal, G. S. Tzelepis, and E. E. Aksoy, “SalsaNext: Fast Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving,” arXiv, 2020.
  5. Y. Zhang et al., “PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation,” in IEEE CVPR, 2020.
  6. C. R. Qi et al., “PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space,” in NeurIPS, 2017.
  7. J. Fei et al., “PillarSegNet: Pillar-Based Semantic Grid Map Estimation Using Sparse LiDAR Data,” in IEEE IV, 2021.
  8. X. Zhu et al., “Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation,” in IEEE CVPR, 2021.
  9. H. Tang et al., “Searching Efficient 3D Architectures with Sparse Point-Voxel Convolution,” in IEEE ECCV, 2020.
  10. J. Xu et al., “RPVNet: A Deep and Efficient Range-Point-Voxel Fusion Network for LiDAR Point Cloud Segmentation,” in IEEE ICCV, 2021.
  11. A. Paszke et al., “ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation,” arXiv, 2016.
  12. F. Piewak et al., “Boosting lidar-based semantic labeling by cross-modal training data generation,” in IEEE ECCV - Workshops, 2018.
  13. R. Cheng et al., “(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Network,” in IEEE CVPR, 2021.
  14. J. Liu et al., “Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awareness,” in NeurIPS, 2020.
  15. J. Zhou and J. Beyerer, “Category Differences Matter: A Broad Analysis of Inter-Category Error in Semantic Segmentation,” in IEEE CVPR - Workshops, 2023.
  16. F. Arnez et al., “A Comparison of Uncertainty Estimation Approaches in Deep Learning Components for Autonomous Vehicle Applications,” in IJCAI - Workshop, 2020.
  17. E. Hüllermeier and W. Waegeman, “Aleatoric and epistemic uncertainty in machine learning: an introduction to concepts and methods,” Springer Machine Learning, 2021.
  18. A. Kendall and Y. Gal, “What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?” in NeurIPS, 2017.
  19. Y. Gal and Z. Ghahramani, “Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning,” in ICML, 2016.
  20. B. Lakshminarayanan, A. Pritzel, and C. Blundell, “Simple and Scalable Predictive Uncertainty Estimation Using Deep Ensembles,” in NeurIPS, 2017.
  21. C. Guo et al., “On Calibration of Modern Neural Networks,” in ICML, 2017.
  22. J. Postels et al., “On the Practicality of Deterministic Epistemic Uncertainty,” in ICML, 2022.
  23. J. Mukhoti et al., “Deep deterministic uncertainty: A new simple baseline,” in IEEE CVPR, 2023.
  24. J. Van Amersfoort et al., “Uncertainty Estimation Using a Single Deep Deterministic Neural Network,” in ICML, 2020.
  25. X. Liu et al., “Severity-aware semantic segmentation with reinforced wasserstein training,” in IEEE CVPR, 2020.
  26. J. Levatić, D. Kocev, and S. Džeroski, “The importance of the label hierarchy in hierarchical multi-label classification,” Journal of Intelligent Information Systems, 2014.
  27. L. Zhang, S. Shah, and I. Kakadiaris, “Hierarchical Multi-label Classification using Fully Associative Ensemble Learning,” Pattern Recognition, 2017.
  28. C. Chen et al., “HAPGN: Hierarchical Attentive Pooling Graph Network for Point Cloud Segmentation,” IEEE Transactions on Multimedia, 2021.
  29. J. Wehrmann, R. Cerri, and R. Barros, “Hierarchical Multi-Label Classification Networks,” in ICML, 2018.
  30. C. Xu and X. Geng, “Hierarchical Classification Based on Label Distribution Learning,” in AAAI, 2019.
  31. E. Giunchiglia and T. Lukasiewicz, “Coherent Hierarchical Multi-Label Classification Networks,” in NeurIPS, 2020.
  32. M. Romero, J. Finke, and C. Rocha, “A top-down supervised learning approach to hierarchical multi-label classification in networks,” Applied Network Science, 2022.
  33. L. Li et al., “Deep hierarchical semantic segmentation,” in IEEE CVPR, 2022.
  34. A. Vaswani, G. Aggarwal, P. Netrapalli, and N. Hegde, “All Mistakes Are Not Equal: Comprehensive Hierarchy Aware Multi-label Predictions (CHAMP),” arXiv, 2022.
  35. T. de Graaff and A. Ribeiro de Menezes, “Capsule networks for hierarchical novelty detection in object classification,” in IEEE IV, 2022.
  36. T.-W. Ke et al., “Unsupervised hierarchical semantic segmentation with multiview cosegmentation and clustering transformers,” in IEEE CVPR, 2022.
  37. E. Ilg et al., “Uncertainty Estimates and Multi-hypotheses Networks for Optical Flow,” in IEEE ECCV, 2018.
  38. C. Sakaridis, D. Dai, and L. Van Gool, “Map-Guided Curriculum Domain Adaptation and Uncertainty-Aware Evaluation for Semantic Nighttime Image Segmentation,” PAMI, 2022.
  39. F. K. Gustafsson, M. Danelljan, and T. B. Schon, “Evaluating scalable bayesian deep learning methods for robust computer vision,” in IEEE CVPR - Workshops, 2020.
  40. M. Dreissig, F. Piewak, and J. Boedecker, “On the calibration of uncertainty estimation in lidar-based semantic segmentation,” in IEEE ITSC, 2023.
  41. D. P. Kingma and J. Ba, “Adam: A Method for Stochastic Optimization,” CoRR, 2014.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets