pith. machine review for the scientific record. sign in

arxiv: 2512.08639 · v3 · submitted 2025-12-09 · 💻 cs.CV · cs.AI

Recognition: no theorem link

Aerial Vision-Language Navigation with a Unified Framework for Spatial, Temporal and Embodied Reasoning

Huilin Xu , Zhuoyang Liu , Yixiang Luomei , Feng Xu
Authors on Pith no claims yet

Pith reviewed 2026-05-16 23:54 UTC · model grok-4.3

classification 💻 cs.CV cs.AI
keywords aerial vision-language navigationmonocular RGBUAV navigationmulti-task learningkeyframe selectionaction mergingnext-token prediction
0
0 comments X

The pith

A model navigates UAVs from egocentric monocular RGB images and language instructions alone by treating navigation as next-token prediction.

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

The paper presents a unified framework for aerial vision-language navigation that operates without depth, panoramic views, or odometry. It casts the task as next-token prediction and jointly trains spatial perception, trajectory reasoning, and action prediction through prompt-guided multi-task learning. Keyframe selection cuts visual redundancy while action merging and label reweighting stabilize training on imbalanced action distributions. On the AerialVLN and OpenFly benchmarks the approach beats prior RGB-only methods in both seen and unseen settings and reduces the gap to stronger RGB-D systems. This design lowers hardware demands for practical UAV deployment in inspection, rescue, and delivery scenarios.

Core claim

The central claim is that a single model using only egocentric monocular RGB observations and natural language instructions can perform aerial VLN by formulating navigation as next-token prediction, jointly optimizing spatial perception, trajectory reasoning, and action prediction via prompt-guided multi-task learning, and employing keyframe selection together with action merging and label reweighting to handle redundancy and supervision imbalance.

What carries the argument

The unified next-token prediction framework that jointly optimizes spatial perception, trajectory reasoning, and action prediction through prompt-guided multi-task learning, augmented by keyframe selection and action merging with label reweighting.

If this is right

  • Navigation succeeds without panoramic images, depth sensors, or odometry on lightweight UAVs.
  • Performance remains competitive in both seen and unseen environments under monocular RGB-only conditions.
  • Prompt-guided multi-task learning enables stable joint optimization of perception, reasoning, and control.
  • Keyframe selection and action merging reduce redundancy and correct long-tailed action distributions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same next-token formulation could be tested on ground robots that must also operate from forward-facing cameras alone.
  • Reducing sensor requirements may allow longer flight times by lowering payload and power draw.
  • The reweighting mechanism might transfer to other long-horizon embodied tasks where action frequencies are uneven.

Load-bearing premise

That egocentric monocular RGB frames contain enough information for spatial, temporal, and embodied reasoning when processed by prompt-guided multi-task learning, keyframe selection, and action merging.

What would settle it

A failure to outperform existing RGB-only baselines on the AerialVLN benchmark in unseen environments would falsify the claim that the unified monocular framework delivers strong results across settings.

Figures

Figures reproduced from arXiv: 2512.08639 by Feng Xu, Huilin Xu, Yixiang Luomei, Zhuoyang Liu.

Figure 1
Figure 1. Figure 1: Aerial vision-language navigation. Left: A drone receives a natural [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Trajectory statistics before and after data preprocessing, showing that [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overview of our framework. Given egocentric keyframes selected from the onboard video stream, our model first encodes the visual observations [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Schematic diagram of the STC module with grid size [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Unified prompting interface for the proposed model. Through task [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Qualitative comparison between predicted and ground-truth drone [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Comparison between OpenFly and our method across three difficulty [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualizations of our method on AerialVLN-S benchmark. Our model successfully follows detailed long-horizon instructions, grounds visual landmarks, [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Qualitative results on auxiliary VQA tasks. Top: examples from the Open3D-VQA dataset [13], where the model answers 3D spatial queries including [PITH_FULL_IMAGE:figures/full_fig_p010_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Qualitative results on trajectory summary task. The model summarizes the route by identifying key landmarks (building, pipe, walkway, garage) [PITH_FULL_IMAGE:figures/full_fig_p011_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Success rate across different trajectory lengths on the AerialVLN-S [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
read the original abstract

Aerial Vision-and-Language Navigation (VLN) aims to enable unmanned aerial vehicles (UAVs) to interpret natural language instructions and navigate complex urban environments using onboard visual observation. This task holds promise for real-world applications such as low-altitude inspection, search-and-rescue, and autonomous aerial delivery. Existing methods often rely on panoramic images, depth inputs, or odometry to support spatial reasoning and action planning. These requirements increase system cost and integration complexity, thus hindering practical deployment for lightweight UAVs. We present a unified aerial VLN framework that operates solely on egocentric monocular RGB observations and natural language instructions. The model formulates navigation as a next-token prediction problem, jointly optimizing spatial perception, trajectory reasoning, and action prediction through prompt-guided multi-task learning. Moreover, we propose a keyframe selection strategy to reduce visual redundancy by retaining semantically informative frames, along with an action merging and label reweighting mechanism that mitigates long-tailed supervision imbalance and facilitates stable multi-task co-training. Extensive experiments on the AerialVLN and OpenFly benchmark validate the effectiveness of our method. Under the challenging monocular RGB-only setting, our model achieves strong results across both seen and unseen environments. It significantly outperforms existing RGB-only baselines and narrows the performance gap with state-of-the-art panoramic RGB-D counterparts. Comprehensive ablation studies further demonstrate the contribution of our task design and architectural choices. Our code is publicly available at https://github.com/return-sleep/AeroAct.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper proposes a unified framework for aerial vision-language navigation (VLN) that operates exclusively on egocentric monocular RGB observations and natural language instructions. Navigation is formulated as next-token prediction and jointly optimized via prompt-guided multi-task learning for spatial perception, trajectory reasoning, and action prediction. Key contributions include a keyframe selection strategy to reduce visual redundancy and an action merging mechanism with label reweighting to address long-tailed supervision. Experiments on the AerialVLN and OpenFly benchmarks report strong results in seen and unseen environments under the RGB-only setting, with claims of outperforming prior RGB-only baselines and narrowing the gap to panoramic RGB-D methods.

Significance. If the reported gains hold under rigorous scrutiny, the work could enable practical deployment of lightweight UAVs for VLN tasks by removing the need for depth, odometry, or panoramic sensors, lowering cost and complexity for applications such as low-altitude inspection and search-and-rescue. Public code release is a positive factor for reproducibility.

major comments (2)
  1. Abstract: the claim that the model 'significantly outperforms existing RGB-only baselines and narrows the performance gap with state-of-the-art panoramic RGB-D counterparts' is presented without any quantitative metrics, success rates, SPL values, error bars, or ablation numbers. This absence prevents assessment of whether the data actually supports the central assertion of robust spatial and embodied reasoning from monocular RGB alone.
  2. Method section (prompt-guided multi-task learning and keyframe/action modules): the framework extracts spatial/embodied cues solely through implicit 2D feature correlations learned from RGB, without explicit depth, 3D reconstruction, or odometry. Given altitude variation and occlusions typical in aerial urban scenes, it remains unclear whether these correlations generalize across unseen environments or remain environment-specific; a concrete test (e.g., cross-altitude or occlusion-specific ablations) is needed to substantiate the embodied-reasoning claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. The comments highlight opportunities to strengthen the presentation of quantitative results and to provide additional evidence for generalization. We address each point below and will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: Abstract: the claim that the model 'significantly outperforms existing RGB-only baselines and narrows the performance gap with state-of-the-art panoramic RGB-D counterparts' is presented without any quantitative metrics, success rates, SPL values, error bars, or ablation numbers. This absence prevents assessment of whether the data actually supports the central assertion of robust spatial and embodied reasoning from monocular RGB alone.

    Authors: We agree that the abstract would benefit from explicit quantitative support. The full manuscript reports success rates, SPL, and comparisons in Tables 1–3 and the ablation studies. We will revise the abstract to include key metrics (e.g., success-rate gains over RGB-only baselines and the remaining gap to RGB-D methods) drawn directly from those results, along with reference to error bars where applicable. revision: yes

  2. Referee: Method section (prompt-guided multi-task learning and keyframe/action modules): the framework extracts spatial/embodied cues solely through implicit 2D feature correlations learned from RGB, without explicit depth, 3D reconstruction, or odometry. Given altitude variation and occlusions typical in aerial urban scenes, it remains unclear whether these correlations generalize across unseen environments or remain environment-specific; a concrete test (e.g., cross-altitude or occlusion-specific ablations) is needed to substantiate the embodied-reasoning claim.

    Authors: The current evaluation already shows competitive performance on unseen environments in both AerialVLN and OpenFly, which contain altitude and occlusion variations. The prompt-guided multi-task objective and keyframe selection are designed to encourage learning of spatial and embodied cues from RGB alone. We nevertheless agree that targeted ablations would provide stronger substantiation and will add cross-altitude and occlusion-specific experiments in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical method evaluated on external benchmarks

full rationale

The paper proposes an empirical aerial VLN framework using prompt-guided multi-task learning, keyframe selection, and action merging on monocular RGB inputs. Navigation is cast as next-token prediction and optimized jointly, with results reported on the external AerialVLN and OpenFly benchmarks. No equations, uniqueness theorems, or self-citations are invoked to derive performance claims; all reported gains are measured against independent baselines and ground-truth trajectories. The central claims therefore rest on experimental outcomes rather than any reduction of predictions to fitted inputs or self-referential definitions.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard domain assumptions from vision-language navigation and multi-task learning; no explicit free parameters, new entities, or ad-hoc axioms are stated in the abstract.

axioms (1)
  • domain assumption Navigation actions and spatial-temporal reasoning can be jointly modeled as next-token prediction in a prompt-guided multi-task setup
    This is the core formulation used to optimize perception, trajectory, and action components together.

pith-pipeline@v0.9.0 · 5575 in / 1194 out tokens · 35042 ms · 2026-05-16T23:54:17.959679+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Vision-and-Language Navigation for UAVs: Progress, Challenges, and a Research Roadmap

    cs.RO 2026-04 unverdicted novelty 4.0

    A survey of UAV vision-and-language navigation that establishes a methodological taxonomy, reviews resources and challenges, and proposes a forward-looking research roadmap.

  2. Vision-Language Navigation for Aerial Robots: Towards the Era of Large Language Models

    cs.RO 2026-04 unverdicted novelty 4.0

    This survey organizes aerial vision-language navigation methods into five architectural categories, critically reviews evaluation infrastructure, and synthesizes seven open problems for LLM/VLM integration.

Reference graph

Works this paper leans on

48 extracted references · 48 canonical work pages · cited by 2 Pith papers · 3 internal anchors

  1. [1]

    Emerging uav technology for disaster detection, mitigation, response, and preparedness,

    A. Khan, S. Gupta, and S. K. Gupta, “Emerging uav technology for disaster detection, mitigation, response, and preparedness,”Journal of Field Robotics, vol. 39, no. 6, pp. 905–955, 2022

    work page 2022

  2. [2]

    Pareto refocusing for drone-view object detection,

    J. Leng, M. Mo, Y . Zhou, C. Gao, W. Li, and X. Gao, “Pareto refocusing for drone-view object detection,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 3, pp. 1320–1334, 2022

    work page 2022

  3. [3]

    High-resolution feature pyramid network for small object detection on drone view,

    Z. Chen, H. Ji, Y . Zhang, Z. Zhu, and Y . Li, “High-resolution feature pyramid network for small object detection on drone view,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 1, pp. 475–489, 2023

    work page 2023

  4. [4]

    Cross-drone transformer network for robust single object tracking,

    G. Chen, P. Zhu, B. Cao, X. Wang, and Q. Hu, “Cross-drone transformer network for robust single object tracking,”IEEE Transactions on Cir- cuits and Systems for Video Technology, vol. 33, no. 9, pp. 4552–4563, 2023

    work page 2023

  5. [5]

    Temporal-spatial feature interac- tion network for multi-drone multi-object tracking,

    H. Wu, H. Sun, K. Ji, and G. Kuang, “Temporal-spatial feature interac- tion network for multi-drone multi-object tracking,”IEEE Transactions on Circuits and Systems for Video Technology, 2024

    work page 2024

  6. [6]

    Aeri- alvln: Vision-and-language navigation for uavs,

    S. Liu, H. Zhang, Y . Qi, P. Wang, Y . Zhang, and Q. Wu, “Aeri- alvln: Vision-and-language navigation for uavs,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 15 384–15 394

    work page 2023

  7. [7]

    Aerial vision-and-dialog navigation,

    Y . Fan, W. Chen, T. Jiang, C. Zhou, Y . Zhang, and X. Wang, “Aerial vision-and-dialog navigation,” inFindings of the Association for Com- putational Linguistics: ACL 2023, 2023, pp. 3043–3061

    work page 2023

  8. [8]

    Citynav: A large-scale dataset for real-world aerial navigation,

    J. Lee, T. Miyanishi, S. Kurita, K. Sakamoto, D. Azuma, Y . Matsuo, and N. Inoue, “Citynav: A large-scale dataset for real-world aerial navigation,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2025, pp. 5912–5922

    work page 2025

  9. [9]

    Towards realistic uav vision-language navigation: Platform, benchmark, and methodology,

    X. Wang, D. Yang, Z. Wang, H. Kwan, J. Chen, W. Wu, H. Li, Y . Liao, and S. Liu, “Towards realistic uav vision-language navigation: Platform, benchmark, and methodology,” inThe Thirteenth International Conference on Learning Representations

    work page

  10. [10]

    Openfly: A versatile toolchain and large-scale benchmark for aerial vision-language navigation,

    Y . Gao, C. Li, Z. You, J. Liu, Z. Li, P. Chen, Q. Chen, Z. Tang, L. Wang, P. Yanget al., “Openfly: A versatile toolchain and large-scale benchmark for aerial vision-language navigation,”arXiv e-prints, pp. arXiv–2502, 2025

    work page 2025

  11. [11]

    Aeroduo: Aerial duo for uav-based vision and language navigation,

    R. Wu, Y . Zhang, J. Chen, L. Huang, S. Zhang, X. Zhou, L. Wang, and S. Liu, “Aeroduo: Aerial duo for uav-based vision and language navigation,” inProceedings of the 33rd ACM International Conference on Multimedia, 2025, pp. 2576–2585

    work page 2025

  12. [12]

    Cityeqa: A hierarchical llm agent on embodied question answering benchmark in city space,

    Y . Zhao, K. Xu, Z. Zhu, Y . Hu, Z. Zheng, Y . Chen, Y . Ji, C. Gao, Y . Li, and J. Huang, “Cityeqa: A hierarchical llm agent on embodied question answering benchmark in city space,”arXiv preprint arXiv:2502.12532, 2025

    work page arXiv 2025

  13. [13]

    Open3dvqa: A benchmark for comprehensive spatial reasoning with multimodal large language model in open space,

    W. Zhang, Z. Zhou, Z. Zheng, C. Gao, J. Cui, Y . Li, X. Chen, and X.-P. Zhang, “Open3dvqa: A benchmark for comprehensive spatial reasoning with multimodal large language model in open space,”arXiv preprint arXiv:2503.11094, 2025

    work page arXiv 2025

  14. [14]

    Aerial vision- and-language navigation via semantic-topo-metric representation guided llm reasoning,

    Y . Gao, Z. Wang, L. Jing, D. Wang, X. Li, and B. Zhao, “Aerial vision- and-language navigation via semantic-topo-metric representation guided llm reasoning,”arXiv preprint arXiv:2410.08500, 2024

    work page arXiv 2024

  15. [15]

    CityNavAgent: Aerial vision-and-language navigation with hierarchical semantic planning and global memory,

    W. Zhang, C. Gao, S. Yu, R. Peng, B. Zhao, Q. Zhang, J. Cui, X. Chen, and Y . Li, “CityNavAgent: Aerial vision-and-language navigation with hierarchical semantic planning and global memory,” inProceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Association for Computational Linguistics, 2025, pp...

    work page 2025

  16. [16]

    Uav-on: A benchmark for open-world object goal navigation with aerial agents,

    J. Xiao, Y . Sun, Y . Shao, B. Gan, R. Liu, Y . Wu, W. Guan, and X. Deng, “Uav-on: A benchmark for open-world object goal navigation with aerial agents,” inProceedings of the 33rd ACM International Conference on Multimedia, 2025, p. 13023–13029

    work page 2025

  17. [17]

    Aerial vision-and-language navigation with grid-based view selection and map construction,

    G. Zhao, G. Li, J. Pan, and Y . Yu, “Aerial vision-and-language navigation with grid-based view selection and map construction,”arXiv preprint arXiv:2503.11091, 2025. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 14

    work page arXiv 2025

  18. [18]

    Flightgpt: Towards generalizable and interpretable uav vision-and-language navigation with vision-language models,

    H. Cai, J. Dong, J. Tan, J. Deng, S. Li, Z. Gao, H. Wang, Z. Su, A. Sumalee, and R. Zhong, “Flightgpt: Towards generalizable and interpretable uav vision-and-language navigation with vision-language models,” inProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 2025, pp. 6670–6687

    work page 2025

  19. [19]

    Vla-an: An efficient and onboard vision-language-action frame- work for aerial navigation in complex environments,

    Y . Wu, M. Zhu, X. Li, Y . Du, Y . Fan, W. Li, Z. Han, X. Zhou, and F. Gao, “Vla-an: An efficient and onboard vision-language-action frame- work for aerial navigation in complex environments,”arXiv preprint arXiv:2512.15258, 2025

    work page arXiv 2025

  20. [20]

    Vision-language navigation: a survey and taxonomy,

    W. Wu, T. Chang, X. Li, Q. Yin, and Y . Hu, “Vision-language navigation: a survey and taxonomy,”Neural Computing and Applications, vol. 36, no. 7, pp. 3291–3316, 2024

    work page 2024

  21. [21]

    Navcomposer: Composing language instructions for navigation trajectories through action-scene-object modularization,

    Z. He, L. Wang, L. Chen, C. Liu, and Q. Chen, “Navcomposer: Composing language instructions for navigation trajectories through action-scene-object modularization,”IEEE Transactions on Circuits and Systems for Video Technology, pp. 1–1, 2025

    work page 2025

  22. [23]

    Room-across- room: Multilingual vision-and-language navigation with dense spa- tiotemporal grounding,

    A. Ku, P. Anderson, R. Patel, E. Ie, and J. Baldridge, “Room-across- room: Multilingual vision-and-language navigation with dense spa- tiotemporal grounding,” inProceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020, pp. 4392–4412

    work page 2020

  23. [24]

    Enhancing vision and language navigation with prompt-based scene knowledge,

    Z. Zhan, J. Qin, W. Zhuo, and G. Tan, “Enhancing vision and language navigation with prompt-based scene knowledge,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 34, no. 10, pp. 9745– 9756, 2024

    work page 2024

  24. [25]

    Beyond the nav-graph: Vision-and-language navigation in continuous environments,

    J. Krantz, E. Wijmans, A. Majumdar, D. Batra, and S. Lee, “Beyond the nav-graph: Vision-and-language navigation in continuous environments,” inEuropean Conference on Computer Vision. Springer, 2020, pp. 104– 120

    work page 2020

  25. [26]

    Vln bert: A recurrent vision-and-language bert for navigation,

    Y . Hong, Q. Wu, Y . Qi, C. Rodriguez-Opazo, and S. Gould, “Vln bert: A recurrent vision-and-language bert for navigation,” inProceedings of the IEEE/CVF conference on Computer Vision and Pattern Recognition, 2021, pp. 1643–1653

    work page 2021

  26. [27]

    Gridmm: Grid memory map for vision-and-language navigation,

    Z. Wang, X. Li, J. Yang, Y . Liu, and S. Jiang, “Gridmm: Grid memory map for vision-and-language navigation,” inProceedings of the IEEE/CVF International conference on computer vision, 2023, pp. 15 625–15 636

    work page 2023

  27. [28]

    Etpnav: Evolving topological planning for vision-language navigation in continuous environments,

    D. An, H. Wang, W. Wang, Z. Wang, Y . Huang, K. He, and L. Wang, “Etpnav: Evolving topological planning for vision-language navigation in continuous environments,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

    work page 2024

  28. [29]

    Dreamwalker: Mental planning for continuous vision-language navigation,

    H. Wang, W. Liang, L. Van Gool, and W. Wang, “Dreamwalker: Mental planning for continuous vision-language navigation,” inProceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 10 873–10 883

    work page 2023

  29. [30]

    Think holistically, act down-to-earth: A semantic navigation strategy with continuous environmental representation and multi-step forward planning,

    B. Chen, J. Kang, P. Zhong, Y . Cui, S. Lu, Y . Liang, and J. Wang, “Think holistically, act down-to-earth: A semantic navigation strategy with continuous environmental representation and multi-step forward planning,”IEEE Transactions on Circuits and Systems for Video Tech- nology, vol. 34, no. 5, pp. 3860–3875, 2024

    work page 2024

  30. [31]

    Navid: Video-based vlm plans the next step for vision- and-language navigation,

    J. Zhang, K. Wang, R. Xu, G. Zhou, Y . Hong, X. Fang, Q. Wu, Z. Zhang, and H. Wang, “Navid: Video-based vlm plans the next step for vision- and-language navigation,”Robotics: Science and Systems, 2024

    work page 2024

  31. [32]

    Navila: Legged robot vision-language- action model for navigation,

    A.-C. Cheng, Y . Ji, Z. Yang, Z. Gongye, X. Zou, J. Kautz, E. Bıyık, H. Yin, S. Liu, and X. Wang, “Navila: Legged robot vision-language- action model for navigation,” inRSS, 2025

    work page 2025

  32. [33]

    Monodream: Monocular vision-language navigation with panoramic dreaming,

    S. Wang, Y . Wang, W. Li, Y . Wang, M. Chen, K. Wang, Z. Su, X. Cai, Y . Jin, D. Liet al., “Monodream: Monocular vision-language navigation with panoramic dreaming,”arXiv preprint arXiv:2508.02549, 2025

    work page arXiv 2025

  33. [34]

    Uni-NaVid: A Video-based Vision-Language-Action Model for Unifying Embodied Navigation Tasks

    J. Zhang, K. Wang, S. Wang, M. Li, H. Liu, S. Wei, Z. Wang, Z. Zhang, and H. Wang, “Uni-navid: A video-based vision-language- action model for unifying embodied navigation tasks,”arXiv preprint arXiv:2412.06224, 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  34. [35]

    Vln-r1: Vision- language navigation via reinforcement fine-tuning,

    Z. Qi, Z. Zhang, Y . Yu, J. Wang, and H. Zhao, “Vln-r1: Vision- language navigation via reinforcement fine-tuning,”arXiv preprint arXiv:2506.17221, 2025

    work page arXiv 2025

  35. [36]

    Reactive navigation of an unmanned aerial vehicle with perception- based obstacle avoidance constraints,

    B. Lindqvist, S. S. Mansouri, J. Halu ˇska, and G. Nikolakopoulos, “Reactive navigation of an unmanned aerial vehicle with perception- based obstacle avoidance constraints,”IEEE Transactions on Control Systems Technology, vol. 30, no. 5, pp. 1847–1862, 2021

    work page 2021

  36. [37]

    Learning vision-based agile flight via differentiable physics,

    Y . Zhang, Y . Hu, Y . Song, D. Zou, and W. Lin, “Learning vision-based agile flight via differentiable physics,”Nature Machine Intelligence, pp. 1–13, 2025

    work page 2025

  37. [38]

    Development of uav- based target tracking and recognition systems,

    S. Wang, F. Jiang, B. Zhang, R. Ma, and Q. Hao, “Development of uav- based target tracking and recognition systems,”IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 8, pp. 3409–3422, 2019

    work page 2019

  38. [39]

    Effective uav navigation for cellular-assisted radio sensing, imaging, and tracking,

    A. V . Savkin, W. Ni, and M. Eskandari, “Effective uav navigation for cellular-assisted radio sensing, imaging, and tracking,”IEEE Transac- tions on Vehicular Technology, vol. 72, no. 10, pp. 13 729–13 733, 2023

    work page 2023

  39. [40]

    Na vid-4d: Unleashing spatial intelligence in egocentric rgb-d videos for vision-and-language navigation,

    H. Liu, W. Wan, X. Yu, M. Li, J. Zhang, B. Zhao, Z. Chen, Z. Wang, Z. Zhang, and H. Wang, “Na vid-4d: Unleashing spatial intelligence in egocentric rgb-d videos for vision-and-language navigation,” in2025 IEEE International Conference on Robotics and Automation (ICRA), 2025, pp. 10 607–10 615

    work page 2025

  40. [41]

    Gqa: A new dataset for real-world visual reasoning and compositional question answering,

    D. A. Hudson and C. D. Manning, “Gqa: A new dataset for real-world visual reasoning and compositional question answering,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 6700–6709

    work page 2019

  41. [42]

    Nvila: Efficient frontier visual language models,

    Z. Liu, L. Zhu, B. Shi, Z. Zhang, Y . Lou, S. Yang, H. Xi, S. Cao, Y . Gu, D. Liet al., “Nvila: Efficient frontier visual language models,” inPro- ceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 4122–4134

    work page 2025

  42. [43]

    Sigmoid loss for language image pre-training,

    X. Zhai, B. Mustafa, A. Kolesnikov, and L. Beyer, “Sigmoid loss for language image pre-training,” inProceedings of the IEEE/CVF international conference on computer vision, 2023, pp. 11 975–11 986

    work page 2023

  43. [44]

    Qwen2 Technical Report

    A. Yang, B. Yang, B. Hui, B. Zheng, B. Yu, C. Zhou, C. Li, C. Li, D. Liu, F. Huang, G. Dong, H. Wei, H. Lin, J. Tang, J. Wang, J. Yang, J. Tu, J. Zhang, J. Ma, J. Xu, J. Zhou, J. Bai, J. He, J. Lin, K. Dang, K. Lu, K. Chen, K. Yang, M. Li, M. Xue, N. Ni, P. Zhang, P. Wang, R. Peng, R. Men, R. Gao, R. Lin, S. Wang, S. Bai, S. Tan, T. Zhu, T. Li, T. Liu, W....

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  44. [45]

    Mapgpt: Map-guided prompting with adaptive path planning for vision- and-language navigation,

    J. Chen, B. Lin, R. Xu, Z. Chai, X. Liang, and K.-Y . K. Wong, “Mapgpt: Map-guided prompting with adaptive path planning for vision- and-language navigation,” inProceedings of the 62nd Annual Meeting of the Association for Computational Linguistics, 2024

    work page 2024

  45. [46]

    GPT-4 Technical Report

    J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F. L. Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkatet al., “Gpt-4 technical report,”arXiv preprint arXiv:2303.08774, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  46. [47]

    Mapping instructions to actions in 3D environments with visual goal prediction,

    D. Misra, A. Bennett, V . Blukis, E. Niklasson, M. Shatkhin, and Y . Artzi, “Mapping instructions to actions in 3D environments with visual goal prediction,” inProceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, E. Riloff, D. Chiang, J. Hockenmaier, and J. Tsujii, Eds. Brussels, Belgium: Association for Computational Li...

    work page 2018

  47. [48]

    Vision-and-language navigation: Interpreting visually-grounded navigation instructions in real environments,

    P. Anderson, Q. Wu, D. Teney, J. Bruce, M. Johnson, N. S ¨underhauf, I. Reid, S. Gould, and A. Van Den Hengel, “Vision-and-language navigation: Interpreting visually-grounded navigation instructions in real environments,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 3674–3683

    work page 2018

  48. [49]

    Awq: Activation-aware weight quanti- zation for on-device llm compression and acceleration,

    J. Lin, J. Tang, H. Tang, S. Yang, W.-M. Chen, W.-C. Wang, G. Xiao, X. Dang, C. Gan, and S. Han, “Awq: Activation-aware weight quanti- zation for on-device llm compression and acceleration,”Proceedings of machine learning and systems, vol. 6, pp. 87–100, 2024. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 15 Huilin Xu(Student Member, IEEE) rec...

    work page 2024