Pith. sign in

REVIEW 4 major objections 5 minor 38 references

UnsDrive: Towards Robust End-to-End Autonomous Driving in Unstructured Scenes

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read UnsDrive claims that explicitly modeling unknown space—alongside occupied and free space—enables safer end-to-end autonomous driving in unstructured open-pit mining scenes, with open-loop collision rates of 0.10% and closed-loop success…

desk verdict A well-built mining-domain driving system whose central claim about unknown-space reasoning is not yet isolated from occupancy conditioning; worth a serious referee with targeted revisions. read the letter →

arxiv 2608.09098 v1 pith:EEOHAKCU submitted 2026-08-10 cs.RO

Nanxin Zeng , Ruiqi Song , Xiangyu Guo , Baiyong Ding , Yunfeng Ai This is my paper · ORCID
classification cs.RO
keywords end-to-endautonomousdrivingunstructuredminingscenesunknown-awareoccupancyflowmatchingclosed-loopsimulationopen-pitpartialobservabilitytrajectoryconsistency
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that end-to-end autonomous driving can work in unstructured open-pit mining scenes, where urban planners struggle because roads lack clear structure and much of the scene is occluded or degraded. Its central claim is that explicitly representing three kinds of space—occupied, free, and unobserved—and conditioning a generative flow-matching planner on that representation yields substantially safer and more accurate driving than adapting urban baselines. The authors support this with UnsDrive, a planner that derives voxel-wise visibility from multi-frame LiDAR, modulates semantic occupancy with that visibility, and adds losses and a scorer that penalize trajectories entering unknown or non-traversable regions. They also build MineLoop, a closed-loop mining simulator, and report that UnsDrive achieves an average trajectory error of 0.44 m and collision rate of 0.10% open-loop, and a driving score of 84.25 with 68.56% route success closed-loop, beating the strongest urban baseline by clear margins. The paper is trying to establish that unknown-space reasoning, not just occupancy geometry, is the load-bearing ingredient for robust planning under partial observability.

What carries the argument

The load-bearing object is the unknown-aware occupancy representation. Visibility per voxel is $V(v) = n_{\mathrm{obs}}(v)/N$, the fraction of $N$ LiDAR frames that observed voxel $v$; the semantic occupancy distribution is rescaled by $V(v)$ and the remaining probability mass is assigned to an explicit unknown class, $P_{\mathrm{unk}}(v) = 1 - V(v)$. This visibility-modulated distribution is fused with learned occupancy features through a channel-wise gate, and the result is vertically compressed and flattened into occupancy tokens. A flow-matching planner (conditional flow matching, CFM) regresses a velocity field toward expert trajectories, with the occupancy tokens attended to first in a safety-first cross-attention cascade; an occupancy trajectory consistency loss and a margin-based scorer then penalize planned waypoints that fall in non-traversable or unknown cells.

What would settle it

Run UnsDrive and the strongest baseline in MineLoop with the visibility channel severed (for example, replacing the LiDAR-derived $V(v)$ with a constant), and check whether the collision-rate and success-rate gaps persist; alternatively, deploy the trained planner on a real mining truck with dust and terrain occlusion and measure whether the closed-loop success advantage over the same baselines holds.

Watch

Extended reading notes

Core claim

UnsDrive's central assertion is that in unstructured mining environments, the unknown is as important as the free and the occupied. The paper constructs an unknown-aware volumetric occupancy representation by ray-tracing multi-frame LiDAR sweeps through a voxel grid, labeling each voxel by how many frames observed it, and assigning the unobserved probability mass to a dedicated unknown class. A conditional flow-matching planner, conditioned first on these occupancy tokens and then on BEV and agent features, generates multimodal trajectories, while an occupancy trajectory consistency loss and an uncertainty-aware scorer ensure that candidate trajectories are ranked down when they enter non-traversable or unobserved space. In the companion MineLoop simulator, this design produces the reported improvements over SparseWorld, DiffusionDrive, OccWorld, and OccNet in both open-loop and closed-loop tests.

Load-bearing premise

Everything is measured in simulation, so the paper's central claim rests on the assumption that MineLoop's simulated mining scenes, visibility degradation, and heavy-vehicle agents reproduce the conditions a real open-pit truck would face closely enough that the measured gains carry over to deployment.

Editorial extensions

If this is right

  • Explicitly modeling unknown space improves both trajectory accuracy and collision avoidance in unstructured scenes, as the ablations show: adding unknown-aware occupancy conditioning, the consistency loss, and the scorer cuts average L2 error by more than half and collision rate by nearly three quarters.
  • Flow-matching generative planning conditioned on safety- and uncertainty-aware representations can outperform diffusion- and occupancy-world baselines even when those baselines are given occupancy supervision.
  • MineLoop provides a closed-loop evaluation harness where mining-specific geometries, visibility degradation, and heavy-vehicle interactions can be tested across loading, hauling, and dumping workflows.
  • The safety benefits compound under closed-loop execution: the success-rate gap over the strongest baseline (about 10 percentage points) is larger than the open-loop error gap alone would suggest, indicating that unknown-aware reasoning improves long-horizon feedback stability.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the visibility-modulated unknown class is the real driver of the gains, a similar design could benefit any off-road or subterranean autonomy setting where occlusion dominates, not just open-pit mines.
  • The reported numbers depend on MineLoop's fidelity; a natural next test is to run UnsDrive on a real open-pit truck with the same sensor stack and see whether the 10-point success-rate advantage survives real dust, vibration, and terrain variation.
  • The paper leaves implicit that the unknown-aware representation could also inform the sensing strategy itself, for example by planning routes that deliberately reduce future unknown volume, a testable extension that the current trajectory-generation formulation does not explore.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. UnsDrive proposes an end-to-end generative planner for autonomous driving in unstructured open-pit mining scenes. The method augments a SurroundOcc-derived volumetric occupancy representation with multi-frame LiDAR-based visibility, creating an unknown-aware occupancy encoding that conditions a conditional flow matching (CFM) planner. An occupancy trajectory consistency (OTC) loss and an uncertainty-aware trajectory scorer are added to penalize trajectories entering non-traversable or unobserved regions. The paper also introduces MineLoop, a closed-loop mining simulator built on the authors' earlier PMWorld platform, and reports open-loop and closed-loop experiments in which UnsDrive outperforms adapted urban-driving baselines, including lower average L2 error and collision rate and higher driving score and success rate. The concluding limitation acknowledges that all experiments are conducted in simulation.

Significance. If the reported results are robust, the paper makes a useful contribution to end-to-end driving under partial observability in unstructured environments: explicit visibility-based unknown-space modeling is a sensible and relatively underexplored idea, the flow-matching planner is well motivated for multimodal mining maneuvers, and the OTC/scorer design directly couples planning with occupancy safety. The MineLoop simulator is potentially valuable as a closed-loop evaluation platform for mining autonomy, assuming its fidelity can be established. The paper also contains strengths worth acknowledging: the visibility estimation in Eq. (1) is parameter-free, the ablation tables show monotonic improvements as components are added, and the experiments cover both open-loop and closed-loop settings. However, the central claim that explicit unknown-space reasoning drives the gains is not yet isolated by the ablations, and the comparison lacks statistical grounding and a fully specified baseline adaptation protocol, so the significance of the headline numbers is currently uncertain.

major comments (4)
  1. [§5.3, Tables 3 and 4] The ablation of UA-OCC does not isolate unknown-awareness from the presence of occupancy conditioning. Eqs. (2)–(4) define UA-OCC as visibility-modulated semantics fused into occupancy tokens; the tables toggle the whole component as a single binary. Without a control row that retains raw SurroundOcc occupancy tokens but removes only the visibility modulation, the reported 26% L2 reduction and one-third collision-rate reduction cannot be attributed to unknown-space reasoning rather than to adding dense occupancy conditioning at all. The same issue affects the OTC ablation: Eq. (7) includes both P_occ and alpha*P_unk, but no row removes only the unknown term, so the specific contribution of penalizing unobserved space is not tested. The authors should add a control with raw occupancy tokens and P_occ-only OTC/scorer to support the paper's central causal claim.
  2. [§5.1–§5.2, Tables 1 and 2] All experimental comparisons are reported without error bars, multiple seeds, or statistical tests. The closed-loop metrics (Driving Score, Success Rate) are aggregate numbers over rollouts, and the open-loop metrics are single runs; the gap between UnsDrive and SparseWorld (0.44 vs. 0.85 m L2, 84.25 vs. 77.36 driving score) could be well within seed-to-seed or episode-to-episode variability. The authors should report means and variances over at least three seeds for open-loop training and over the full set of MineLoop benchmark routes for closed-loop evaluation, with significance tests where appropriate. This is load-bearing because the paper's central conclusion is an empirical superiority claim.
  3. [§5.1–§5.2, baseline comparison] The adaptation protocol for the baselines (OccWorld, OccNet, SparseWorld, DiffusionDrive) is not described. It is unclear which input modalities each baseline received, whether they were trained on the same MineLoop-collected dataset or on their own pretraining data, what training schedules and hyperparameters were used, and how their outputs were converted into the same trajectory format for open- and closed-loop evaluation. Without this information, the fairness of the comparison cannot be assessed, and the large margin over SparseWorld may partly reflect differences in training effort or input conditioning rather than the proposed unknown-aware representation.
  4. [§4 and §6] The claim that UnsDrive improves robustness for real open-pit mining rests on the fidelity of MineLoop, which is built by the same group and extends the authors' earlier PMWorld simulator. The paper provides no validation of MineLoop's visibility degradation model, terrain occlusion, heavy-vehicle dynamics, or reactive agent behavior against real mining-site data, and the conclusion explicitly states that all experiments are in simulation. Given that the reported success-rate gain (68.56% vs. 58.27%) could change under different visibility or agent models, the authors should provide at least a sensitivity analysis over MineLoop's key simulation parameters (dust levels, sensor dropout, agent density) or a comparison against real sensor logs to substantiate the transferability claim.
minor comments (5)
  1. [Figure 2] The figure contains stray question-mark symbols and an unclear 'τt encoder' label; these should be cleaned up or replaced with proper notation.
  2. [§3.2, Eq. (2)] The visibility-modulated semantics should explicitly state that the original semantic distribution p_raw is renormalized or that P_unk absorbs all remaining probability mass; a short clarification would prevent ambiguity about the total probability mass of the resulting distribution.
  3. [§3.2, Eq. (1)] The paper should specify the voxel resolution and the number N of LiDAR frames used for visibility estimation, since these choices directly affect the occupancy tokenization and the downstream planner's perception of unknown regions.
  4. [§5.1, evaluation metrics] The definitions of 'off-road behavior' and 'prolonged off-road behavior' in the closed-loop termination condition are not given; a precise threshold or definition would make the Success Rate and Driving Score metrics reproducible.
  5. [§5.2, Table 1] The column header 'Aux. Sup.' is abbreviated without explanation; it should be spelled out or defined in the caption, and the table would benefit from a note on how each baseline's auxiliary supervision was obtained in the mining dataset.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation chain; mild self-referential evaluation via in-house MineLoop and PMWorld self-citation.

full rationale

UnsDrive's derivation chain is constructive rather than circular. Equations (1)-(4) define visibility, visibility-modulated semantics, gating, and occupancy tokens; Eq. (5) is the CFM objective; Eqs. (7)-(9) define the OTC regularizer and the scoring objective. None of these is obtained by substituting its own target: P_unk = 1 - V(v) is a modeling definition of unknown from LiDAR visibility, not a prediction of something already assumed; OTC penalizes trajectories by the model's own occupancy output, which is a self-supervised auxiliary objective, not a fitted parameter renamed as a prediction. The main self-referential element is evaluation-side: MineLoop follows the scenario-engineering pipeline and 16-DoF dynamics of PMWorld [2], whose authorship overlaps with this paper, and all closed-loop results are generated inside MineLoop. This creates mild self-referential pressure, but the baselines (OccWorld, OccNet, SparseWorld, DiffusionDrive) are external methods and the reported gains are empirical measurements, so the argument does not reduce to a self-citation. Table 3 toggles UA-OCC as one binary component and never isolates unknown-awareness from the presence of occupancy conditioning; that is an experimental identifiability limitation, not a logical circularity. The Conclusion explicitly states that all experiments are in simulation, which is an external-validity limitation. Overall, no load-bearing circular step exists; the score reflects the in-house simulator and self-cited PMWorld basis rather than any equation-level circularity.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The ledger captures the hand-chosen constants and domain assumptions the results depend on. No invented physical entities are introduced; unknown-aware occupancy is a representational label, not a postulated mechanism. The most consequential entries are the MineLoop fidelity assumption and the LiDAR-visibility ground-truth assumption.

free parameters (6)
  • alpha (unknown-region penalty weight) = unspecified, stated to be in (0,1)
    Appears in Eq. 7 and controls how strongly unknown regions are penalized in the OTC loss; central to the claimed safety benefit.
  • epsilon (safety tolerance) = unspecified
    Appears in Eq. 8 as the hinge threshold in the occupancy trajectory consistency loss.
  • m (ranking margin) = unspecified
    Appears in Eq. 9 in the scorer ranking objective and determines how much separation is required between candidate scores.
  • loss weights lambda_occ, lambda_OTC, lambda_score = unspecified
    Appear in Eq. 10 and balance occupancy learning, consistency regularization, and scorer ranking in the total training objective.
  • K (number of sampled candidate trajectories) = unspecified
    Controls the multimodality and safety of final plan selection; no value is reported.
  • N (number of LiDAR frames for visibility) = unspecified
    Appears in Eq. 1 for the visibility score; affects how much of the scene is labeled observed versus unknown.
assumptions (6)
  • domain assumption LiDAR ray tracing reliably labels occupied, free, and unknown voxels even under dust, rain, and degraded visibility.
    Section 3.2 derives the visibility score V(v) from LiDAR sweeps; if LiDAR returns are unreliable in degraded conditions, the unknown labels are wrong.
  • domain assumption Simulator-generated occupancy supervision and visibility labels are engine-consistent and accurate enough to train and evaluate planning.
    Section 4.3 states MineLoop generates occupancy labels from simulator geometry and agent states; this grounds the occupancy branch of the model.
  • domain assumption MineLoop's virtual scenes and reactive heavy-vehicle agents are representative of real open-pit mining, so closed-loop gains transfer to deployment.
    The conclusion admits all experiments are in simulation; the external validity of the whole evaluation depends on this premise.
  • domain assumption TransFuser and SurroundOcc pretrained representations transfer to unstructured mining scenes after fine-tuning.
    Section 3.2 uses TransFuser for BEV and agent features and SurroundOcc for volumetric occupancy; no mining-domain pretraining is described.
  • standard math Conditional flow matching with linear interpolation is a valid generative objective for multimodal trajectory planning.
    Section 3.3 relies on the CFM formalism of Lipman et al. [17]; this is a standard result from prior literature.
  • domain assumption The four baselines are adapted to the mining task fairly and without disadvantage.
    Section 5.2 compares against OccWorld, OccNet, SparseWorld, and DiffusionDrive, but the adaptation procedure is not described, so fairness of the comparison is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UnsDrive: Towards Robust End-to-End Autonomous Driving in Unstructured Scenes." pith.science (2026). https://pith.science/paper/EEOHAKCU

@misc{pith2026260809098,
  author       = {Pith},
  title        = {Pith review of: UnsDrive: Towards Robust End-to-End Autonomous Driving in Unstructured Scenes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EEOHAKCU}},
  note         = {Machine review of arXiv:2608.09098}
}
read the original abstract

End-to-end planning has shown strong promise for autonomous driving, but most existing methods are designed for structured urban roads and generalize poorly to unstructured mining environments. In such settings, weak road structure, terrain-induced occlusions, degraded visibility, and large unobserved regions make safe planning particularly challenging. To address these challenges, we propose UnsDrive, an end-to-end planner designed for unstructured mining scenes. UnsDrive builds an unknown-aware occupancy representation that explicitly models occupied, free, and unknown space using multi-frame visibility cues, and conditions a flow-matching planner on this representation to generate multimodal future trajectories. To improve safety under partial observability, we further introduce an occupancy trajectory consistency loss and an uncertainty-aware trajectory scorer that penalize trajectories entering non-traversable or unobserved regions. We also present MineLoop, a mining-oriented closed-loop simulator for evaluating autonomous driving under irregular road geometry, degraded visibility, heavy-vehicle interactions, and mining-specific operational constraints. Experiments in both open-loop and closed-loop settings show that UnsDrive consistently outperforms strong baselines in trajectory accuracy, collision avoidance, and long-horizon driving robustness. These results demonstrate the value of explicit unknown-space reasoning for autonomous driving in unstructured mining environments.

Figures

Figures reproduced from arXiv: 2608.09098 by the authors.

Figure 1
Figure 1. Closed-loop pipeline of UnsDrive in MineLoop. MineLoop provides synchronized camera, LiDAR, and ego-state [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of UnsDrive. Multi-view images and LiDAR are encoded into BEV, agent, and occupancy features. An [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. MineLoop scenarios. Representative scenes under different mining conditions. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative results in simulation. Temporal trajectory visualizations in two representative mining scenarios. For [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 31 canonical work pages

  1. [1]

    Joonwoo Ahn, Minsoo Kim, and Jaeheung Park. 2022. Vision-based autonomous driving for unstructured environments using imitation learning.arXiv preprint arXiv:2202.10002(2022)

    work page Pith review arXiv 2022

  2. [2]

    Yunfeng Ai, Yuhang Liu, Yu Gao, Chen Zhao, Xiang Cheng, Jinpeng Han, Bin Tian, Long Chen, and Fei-Yue Wang. 2023. PMWorld: A parallel testing platform for autonomous driving in mines.IEEE Transactions on Intelligent Vehicles9, 1 (2023), 1402–1411

    work page 2023

  3. [3]

    Shahin Atakishiyev, Mohammad Salameh, Hengshuai Yao, and Randy Goebel

    work page

  4. [4]

    Kashyap Chitta, Aditya Prakash, Bernhard Jaeger, Zehao Yu, Katrin Renz, and Andreas Geiger. 2022. Transfuser: Imitation with transformer-based sensor fusion for autonomous driving.IEEE transactions on pattern analysis and machine intelligence45, 11 (2022), 12878–12895

    work page 2022

  5. [5]

    Chenxu Dang, Haiyan Liu, Jason Bao, Pei An, Xinyue Tang, An Pan, Jie Ma, Bingchuan Sun, and Yan Wang. 2026. Sparseworld: A flexible, adaptive, and efficient 4d occupancy world model powered by sparse and dynamic queries. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 40. 3497–3505

    work page 2026

  6. [6]

    Daniel Dauner, Marcel Hallgarten, Tianyu Li, Xinshuo Weng, Zhiyu Huang, Zetong Yang, Hongyang Li, Igor Gilitschenski, Boris Ivanovic, Marco Pavone, et al. 2024. Navsim: Data-driven non-reactive autonomous vehicle simulation and benchmarking.Advances in Neural Information Processing Systems37 (2024), 28706–28719

    work page 2024

  7. [7]

    Wei Dong, Sikai Lu, Xinhe Chen, Shunyao Zhang, Qingchao Liu, Ze Liu, Long Chen, Hai Wang, and Yingfeng Cai. 2025. End-to-End Autonomous Driving: From Classic Paradigm to Large Model Empowerment—A Comprehensive Survey. IEEE Internet of Things Journal13, 3 (2025), 3870–3898

    work page 2025

  8. [8]

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Antonio Lopez, and Vladlen Koltun. 2017. CARLA: An open urban driving simulator. InConference on robot learning. PMLR, 1–16

    2017

Show all 38 references
  1. [9]

    Yuqing Guo, Zelin Guo, Yazhou Wang, Danya Yao, Bai Li, and Li Li. 2023. A survey of trajectory planning methods for autonomous driving—Part I: Unstructured scenarios.IEEE Transactions on Intelligent Vehicles9, 9 (2023), 5407–5434

    2023

  2. [10]

    Atoosa Haghighizadeh, Omid Rajabi, Arman Nezarat, Zahra Hajyani, Mina Hagh- mohammadi, Soheila Hedayatikhah, Soheila Delnabi Asl, and Ali Aghababai Beni

  3. [11]

    Xiaosong Jia, Zhenjie Yang, Qifeng Li, Zhiyuan Zhang, and Junchi Yan. 2024. Bench2drive: Towards multi-ability benchmarking of closed-loop end-to-end autonomous driving.Advances in Neural Information Processing Systems37 (2024), 819–844

    2024

  4. [12]

    Comprehensive analysis of heavy metal soil contamination in mining Envi- ronments: Impacts, monitoring Techniques, and remediation strategies.Arabian Journal of Chemistry17, 6 (2024), 105777

    2024

  5. [13]

    Bo Jiang, Shaoyu Chen, Qing Xu, Bencheng Liao, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, Chang Huang, and Xinggang Wang. 2023. Vad: Vectorized scene representation for efficient autonomous driving. InProceedings of the IEEE/CVF International Conference on Computer Visio...

    2023

  6. [14]

    Xiaosong Jia, Junqi You, Zhiyuan Zhang, and Junchi Yan. 2025. Drivetransformer: Unified transformer for scalable end-to-end autonomous driving.arXiv preprint arXiv:2503.07656(2025)

    2025 arXiv

  7. [15]

    Xincheng Li, Zhaoyi Wang, Yanjun Huang, and Hong Chen. 2023. A survey on self-evolving autonomous driving: a perspective on data closed-loop technology. IEEE Transactions on Intelligent Vehicles8, 11 (2023), 4613–4631

    2023

  8. [16]

    Han Li, Peng Chen, Guizhen Yu, Bin Zhou, Yiming Li, and Yaping Liao. 2023. Trajectory planning for autonomous driving in unstructured scenarios based on deep learning and quadratic optimization.IEEE Transactions on Vehicular Technology73, 4 (2023), 4886–4903

    2023

  9. [17]

    Yaron Lipman, Ricky TQ Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le

  10. [18]

    Bencheng Liao, Shaoyu Chen, Haoran Yin, Bo Jiang, Cheng Wang, Sixu Yan, Xinbang Zhang, Xiangyu Li, Ying Zhang, Qian Zhang, et al. 2025. Diffusiondrive: Truncated diffusion model for end-to-end autonomous driving. InProceedings of the Computer Vision and Pattern Recognition Con...

    2025

  11. [19]

    Marya Rasib, Muhammad Atif Butt, Faisal Riaz, Adel Sulaiman, and Muhammad Akram. 2021. Pixel level segmentation based drivable road region detection and steering angle estimation method for autonomous driving on unstructured roads. IEEE Access9 (2021), 167855–167867

    2021

  12. [20]

    Hao Shao, Letian Wang, Ruobing Chen, Steven L Waslander, Hongsheng Li, and Yu Liu. 2023. Reasonnet: End-to-end driving with temporal and global reasoning. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition. 13723–13733

    2023

  13. [21]

    Chen Min, Shubin Si, Xu Wang, Hanzhang Xue, Weizhong Jiang, Zitong Chen, Mengmeng Li, Jilin Mei, Erke Shang, Zhipeng Xiao, et al . 2024. Autonomous driving in unstructured environments: How far have we come?arXiv preprint arXiv:2410.07701(2024)

    2024

  14. [22]

    Wenwen Tong, Chonghao Sima, Tai Wang, Li Chen, Silei Wu, Hanming Deng, Yi Gu, Lewei Lu, Ping Luo, Dahua Lin, et al. 2023. Scene as occupancy. InProceedings of the IEEE/CVF International Conference on Computer Vision. 8406–8415

    2023

  15. [23]

    Nan Wang, Xiang Li, Kanghua Zhang, Jixin Wang, and Dongxuan Xie. 2024. A survey on path planning for autonomous ground vehicles in unstructured environments.Machines12, 1 (2024), 31

    2024

  16. [24]

    Wenchao Sun, Xuewu Lin, Yining Shi, Chuang Zhang, Haoran Wu, and Sifa Zheng. 2025. Sparsedrive: End-to-end autonomous driving via sparse scene representation. In2025 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 8795–8801

    2025

  17. [25]

    Shuo Xing, Chengyuan Qian, Yuping Wang, Hongyuan Hua, Kexin Tian, Yang Zhou, and Zhengzhong Tu. 2025. Openemma: Open-source multimodal model for end-to-end autonomous driving. InProceedings of the Winter Conference on Applications of Computer Vision. 1001–1009

    2025

  18. [26]

    Zebin Xing, Xingyu Zhang, Yang Hu, Bo Jiang, Tong He, Qian Zhang, Xiaoxiao Long, and Wei Yin. 2025. Goalflow: Goal-driven flow matching for multimodal trajectories generation in end-to-end autonomous driving. In2025 IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2025

  19. [27]

    Yi Wei, Linqing Zhao, Wenzhao Zheng, Zheng Zhu, Jie Zhou, and Jiwen Lu

  20. [28]

    Ekim Yurtsever, Jacob Lambert, Alexander Carballo, and Kazuya Takeda. 2020. A survey of autonomous driving: Common practices and emerging technologies. IEEE access8 (2020), 58443–58469

    2020

  21. [29]

    Jingyuan Zhao, Yuyan Wu, Rui Deng, Susu Xu, Jinpeng Gao, and Andrew Burke

  22. [30]

    Wenzhao Zheng, Weiliang Chen, Yuanhui Huang, Borui Zhang, Yueqi Duan, and Jiwen Lu. 2024. Occworld: Learning a 3d occupancy world model for autonomous driving. InEuropean conference on computer vision. Springer, 55–72

    2024

  23. [31]

    Lu Xiong, Zhiqiang Fu, Dequan Zeng, and Bo Leng. 2021. An optimized tra- jectory planner and motion controller framework for autonomous driving in unstructured environments.Sensors21, 13 (2021), 4409

    2021

  24. [32]

    Runwen Zhu, Jianbo Zhao, Diankun Zhang, Guoan Wang, Xiwu Chen, Siyu Zhang, Jiahao Gong, Qibin Zhou, Wenyuan Zhang, Ningzi Wang, et al . 2025. Sparsead: Sparse query-centric paradigm for efficient end-to-end autonomous driving.IEEE Transactions on Artificial Intelligence(2025)

    2025

  25. [33]

    Zeyu Zhu, Nan Li, Ruoyu Sun, Donghao Xu, and Huijing Zhao. 2020. Off-road autonomous vehicles traversability analysis and trajectory planning based on deep inverse reinforcement learning. In2020 IEEE intelligent vehicles symposium (IV). IEEE, 971–977

    2020

  26. [36]

    Xingcheng Zhou, Xuyuan Han, Feng Yang, Yunpu Ma, Volker Tresp, and Alois Knoll. 2026. Opendrivevla: Towards end-to-end autonomous driving with large vision language action model. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 40. 13782–13790

    2026

  27. [2022]

    Flow matching for generative modeling.arXiv preprint arXiv:2210.02747 (2022)

    2022 arXiv

  28. [2023]

    InProceedings of the IEEE/CVF International Conference on Computer Vision

    Surroundocc: Multi-camera 3d occupancy prediction for autonomous driving. InProceedings of the IEEE/CVF International Conference on Computer Vision. 21729–21740

  29. [2024]

    Explainable artificial intelligence for autonomous driving: A comprehensive overview and field guide for future research directions.IEEE Access12 (2024), 101603–101625

    2024

  30. [2025]

    Surveys57, 10 (2025), 1–60

    A survey of autonomous driving from a deep learning perspective.Comput. Surveys57, 10 (2025), 1–60

    2025

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.