Perceptual Incompleteness in Autonomous Driving: From Information Deficiency to System-Level Propagation
DOI:
https://doi.org/10.54097/wwz9hp26Keywords:
Perceptual Incompleteness, Uncertainty Propagation, Modular Architecture, Autonomous DrivingAbstract
Autonomous driving perception systems face inherent observational incompleteness due to sensor field-of-view constraints and mutual occlusion among targets. The prevailing supervised learning paradigm, which presupposes fully annotated data, fails to provide effective behavioral constraints for models in information-deficient regions. From an information-theoretic perspective, this paper defines perceptual incompleteness as an intrinsic limit imposed jointly by observational geometry and physical mechanisms, and accordingly establishes a three-fold taxonomy comprising spatial, temporal, and causal deficiency. For each type of deficiency, the paper examines its origins, essential characteristics, and existing mitigation strategies. On this basis, the paper examines the systemic limitations of existing approaches from four perspectives: causal coupling, interface attenuation, semantic deviation, and evaluation blind spots. The analysis reveals that the three types of deficiency amplify each other through physical coupling; the perception-to-prediction and prediction-to-planning interfaces filter out uncertainty information through probability binarization and tail branch truncation, respectively; there is no direct mathematical correspondence between perceptual confidence and planning risk probabilities; and current evaluation frameworks systematically exclude information deficiency from the assessment scope. To address these limitations, this paper proposes four improvement strategies: constructing a joint modeling framework to address the causal coupling among the three types of deficiency, designing probabilistic module interfaces to prevent information attenuation, establishing standardization of cross-module probabilistic semantics, and developing a closed-loop risk-oriented evaluation system. The four improvements target training, interface transmission, semantic interpretation, and performance validation respectively, forming a complete solution from algorithm design to system integration to effect validation.
Downloads
References
[1] Hashempoor, H., & Hwang, Y. D. (2025). FastTracker: Real-time and accurate visual tracking [Preprint]. arXiv. https:// arxiv. org/abs/2508.14370.
[2] Rothenhäusler, A., Mazzola, M., Look, A., et al. (2026). Don't double it: Efficient agent prediction in occlusions [Preprint]. arXiv. https://arxiv.org/abs/2601.21504.
[3] Jimenez-Bermejo, V., Trentin, V., Artunedo, A., et al. (2025). Evidential semantic lane-grid for unknown space analysis and high-level representation of dynamic occupancy grids. IEEE Transactions on Intelligent Vehicles, 10(7), 3993–4009. https:// doi. org/10.1109/TIV.2025.3482611.
[4] Chen, X., Gu, G., & Ning, X. (2026). GMS: Graph-guided multi-modal spatio-temporal fusion model for trajectory prediction in autonomous driving. Neurocomputing, 661, 132010. https://doi.org/10.1016/j.neucom.2026.132010.
[5] Ye, H., Xiao, Y., Lu, C., et al. (2026). Vec-QMDP: Vectorized POMDP planning on CPUs for real-time autonomous driving [Preprint]. arXiv. https://arxiv.org/abs/2602.08334.
[6] Yang, Y., Mei, J., Ma, Y., et al. (2025). Driving in the occupancy world: Vision-centric 4D occupancy forecasting and planning via world models for autonomous driving. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 39, No. 9, pp. 9327–9335). AAAI Press. https://doi.org/ 10. 1609/aaai.v39i9.32947.
[7] Shi, Y., Jiang, K., Li, J., et al. (2025). Grid-centric traffic scenario perception for autonomous driving: A comprehensive review. IEEE Transactions on Neural Networks and Learning Systems, 36(7), 11814–11834. https://doi.org/ 10.1109/ TNNLS. 2024.3429678.
[8] Zhang, Y., Zhang, J., Wang, Z., et al. (2026). Vision-based 3D occupancy prediction in autonomous driving: A review and outlook. Frontiers of Computer Science, 20(1), 13–34. https:// doi. org/10.1007/s11704-025-4112-8.
[9] Ping, P., Zhang, X., Tao, L., et al. (2026). A comprehensive survey on multi-sensor information processing and fusion for BEV perception in autonomous vehicles. Information Fusion, 126. https://doi.org/10.1016/j.inffus.2026.103487.
[10] Kaelbling, L. P., Littman, M. L., & Cassandra, A. R. (1998). Planning and acting in partially observable stochastic domains. Artificial Intelligence, 101(1–2), 99–134. https://doi. org/10. 1016/ S0004-3702(98)00023-X.
[11] Kurniawati, H. (2022). Partially observable Markov decision processes and robotics. Annual Review of Control, Robotics, and Autonomous Systems, 5(1), 253–277. https://doi. org/ 10. 1146/ annurev-control-042920-091756.
[12] Hu, Y., Yang, J., Chen, L., et al. (2023). Planning-oriented autonomous driving. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 17853–17862). IEEE. https://doi.org/10. 1109/CVPR 52729. 2023. 01712.
[13] Xu, M., Liu, Z., Wang, B., et al. (2025). A survey of autonomous driving trajectory prediction: Methodologies, challenges, and future prospects. Machines, 13(9), 818. https:// doi. org/10.3390/machines13090818.
[14] Tang, Y., & Ma, W. (2026). INTENT: Trajectory prediction framework with intention-guided contrastive clustering. IEEE Open Journal of Intelligent Transportation Systems, 7, 337–352. https://doi.org/10.1109/OJITS.2026.3582164.
[15] Kendall, A., & Gal, Y. (2017). What uncertainties do we need in Bayesian deep learning for computer vision? In Advances in Neural Information Processing Systems (Vol. 30, pp. 5580–5590). Curran Associates.
[16] Galvão, L. G., & Huda, M. N. (2024). Pedestrian and vehicle behaviour prediction in autonomous vehicle system — A review. Expert Systems with Applications, 238. https://doi.org/ 10. 1016/j.eswa.2024.121892.
[17] Liang, M., Yang, B., Zeng, W., et al. (2020). PnPNet: End-to-end perception and prediction with tracking in the loop. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 11550–11559). IEEE. https:// doi. org/ 10.1109/CVPR42600.2020.01157.
[18] Pang, Z., Ramanan, D., Li, M., et al. (2023). Streaming motion forecasting for autonomous driving. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 7407–7414). IEEE. https://doi.org/ 10.1109/ IRO S55552.2023.10342110.
[19] Papais, S., Wang, L., Jain, M., Rezaei, B., & Waslander, S. L. (2026). BeyondSight: Object permanence for end-to-end autonomous driving. In European Conference on Computer Vision (ECCV). Springer.
[20] Lu, D., Du, H., Wu, Z., et al. (2025). Risk assessment in autonomous driving: A comprehensive survey of risk sources, methodologies, and system architectures. Autonomous Intelligent Systems, 5(1), 103–123. https://doi.org/10. 1007/ s43684-024-00072-9.
[21] Dal'Col, L., Oliveira, M., & Santos, V. (2025). Joint perception and prediction for autonomous driving: A survey. IEEE Transactions on Intelligent Transportation Systems, 26(12), 21427–21452. https://doi.org/10.1109/TITS.2025.3487124.
[22] Guo, C., Pleiss, G., Sun, Y., & Weinberger, K. Q. (2017). On calibration of modern neural networks. In International Conference on Machine Learning (ICML) (pp. 2130–2143). PMLR.
[23] Yang, P., Zheng, Y., Zhang, Q., et al. (2025). UncAD: Towards safe end-to-end autonomous driving via online map uncertainty. In 2025 IEEE International Conference on Robotics and Automation (ICRA) (pp. 6408–6415). IEEE. https://doi.org/ 10. 1109/ ICRA64210.2025.11009278.
[24] Caesar, H., Bankiti, V., Lang, A. H., et al. (2020). nuScenes: A multimodal dataset for autonomous driving. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 11618–11628). IEEE. https:// doi. org/ 10.1109/CVPR42600.2020.01164.
[25] Wei, Y., Zhao, L., Zheng, W., et al. (2023). SurroundOcc: Multi-camera 3D occupancy prediction for autonomous driving. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV) (pp. 21672–21683). IEEE. https://doi. org/ 10.1109/ICCV51070.2023.01984.
[26] Wang, X., Zhu, Z., Xu, W, et al. (2023). OpenOccupancy: A large scale benchmark for surrounding semantic occupancy perception. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV) (pp. 17804–17813). IEEE. https:// doi. org/10.1109/ICCV51070.2023.01638.
[27] Sun, P., Kretzschmar, H., Dotiwalla, X., et al. (2020). Scalability in perception for autonomous driving: Waymo Open Dataset. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 2443–2451). IEEE. https://doi.org/10.1109/CVPR42600.2020.00252.
[28] Jiang, A., Sun, H., Zhao, H., et al. (2026). UniUncer: Unified dynamic static uncertainty for end to end driving [Preprint]. arXiv. https://arxiv.org/abs/2603.07686.
[29] Ivanovic, B., Lin, Y., Shrivastava, S., et al. (2022). Propagating state uncertainty through trajectory forecasting. In 2022 IEEE International Conference on Robotics and Automation (ICRA) (pp. 2351–2358). IEEE. https://doi.org/ 10. 1109/ ICRA 46639. 2022. 9811647.
[30] Deglurkar, S., Shen, H., Muthali, A., et al. (2024). System-level analysis of module uncertainty quantification in the autonomy pipeline. In 2024 IEEE 63rd Conference on Decision and Control (CDC) (pp. 1103–1109). IEEE. https://doi. org/ 10. 1109/ CDC54075.2024.10835946.
[31] Araujo, B., Teixeira, J. F., Fonseca, J., et al. (2024). The road to safety: A review of uncertainty and applications to autonomous driving perception. Entropy, 26(8), 634. https:// doi. org/10.3390/e26080634.
[32] Abdar, M., Pourpanah, F., Hussain, S., et al. (2021). A review of uncertainty quantification in deep learning: Techniques, applications and challenges. Information Fusion, 76, 243–297. https:// doi.org/10.1016/j.inffus.2021.06.008.
[33] Meding, I., Bodin, A., Tonderski, A., et al. (2023). You can have your ensemble and run it too — Deep ensembles spread over time. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV) (pp. 4022–4031). IEEE. https://doi. org/ 10.1109/ICCV51070.2023.00374.
[34] Gao, J., Chen, M., Xiang, L., et al. (2026). A comprehensive survey on evidential deep learning and its applications. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(3), 2118–2138. https://doi.org/10. 1109/ TPAMI. 2025. 352 4112.
[35] Ryu, W., Yu, S., Moon, S., et al. (2025). SUPER-AD: Semantic uncertainty-aware planning for end-to-end robust autonomous driving [Preprint]. arXiv. https://arxiv.org/abs/2511.22865.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Frontiers in Computing and Intelligent Systems
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
