Pith. sign in

REVIEW 3 major objections 4 minor 47 references

Towards Autonomous UAV Visual Object Search in City Space: Benchmark and Agentic Methodology

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper claims that CityAVOS is the first benchmark for autonomous UAV visual object search in urban spaces, and that its PRPSearcher agent reaches 53.50% success rate, far above baselines.

desk verdict A genuinely useful first benchmark for instruction-free aerial visual object search, but the headline SR/SPL numbers rest on self-declared success and a misdefined SPL until the evaluation protocol is fixed. read the letter →

arxiv 2505.08765 v2 pith:Q6CGFPVR submitted 2025-05-13 cs.CV cs.AI

Yatai Ji , Zhengqiu Zhu , Yong Zhao , Beidan Liu , Chen Gao , Yihao Zhao , Sihang Qiu , Yue Hu
show 2 more authors
Quanjun Yin Yong Li
This is my paper · ORCID
classification cs.CVcs.AI
keywords AerialVisualObjectSearchUAVnavigationmulti-modallargelanguagemodelbenchmarkdatasetsemanticmappingexploration-exploitationbalanceurbanembodiedintelligencespatialreasoning
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 introduces CityAVOS, the first benchmark for Aerial Visual Object Search (AVOS) in city spaces, containing 2,420 tasks over six object categories and three difficulty levels. To solve these tasks, it proposes PRPSearcher, an agent built on multimodal large language models that constructs three maps: object-centric semantic, attraction-based cognitive, and uncertainty maps, plus an IPT prompting mechanism to balance exploration and exploitation. On a 605-task subset, PRPSearcher achieves 53.50% success rate and 40.57% SPL, substantially exceeding the compared baselines. The reported gap to human performance (78.68% SR) indicates that semantic reasoning and spatial exploration in AVOS remain open problems.

What carries the argument

The load-bearing mechanism is the triple-map architecture carried through the agent's loop. The object-centric 3D dynamic semantic map restricts semantic segmentation to objects the MLLM judges relevant to the target, keeping the map sparse. The 3D cognitive map stores an attraction value $A(s)\in[0,1]$ per semantic category, so high-attraction clusters (found by DBSCAN) mark likely target locations; a mirrored map zeroes out already-recognized grids, which is the denoising step. The 3D uncertainty map attenuates each cell's uncertainty by $f(d)=e^{-\alpha d}$ with distance, so the exploration advice is the action maximizing total uncertainty reduction. The Inspiration Promote Thought (IPT) prompt injects that exploration advice only when the reward exceeds threshold $\theta$, letting the MLLM planner combine long-term exploitation guidance with occasional 'inspiration' to explore.

What would settle it

Re-run PRPSearcher on the same 605-task subset and have an independent verifier compare the object at the stop location with the ground-truth target; if the verified success rate is substantially below the reported 53.50%, the headline result is inflated by self-reported stops.

Watch

Extended reading notes

Core claim

The central claim is that autonomous UAV search for common urban objects, given only an image and text description, can be formalized as a benchmark task and solved better by an MLLM-driven agent than by existing navigation methods. CityAVOS provides the standardized tasks, and PRPSearcher demonstrates the approach: it selectively segments only target-related semantics into a 3D dynamic semantic map, assigns MLLM-derived attraction values to build a 3D cognitive map, tracks exploration with a 3D uncertainty map, and uses thresholded exploration advice in its prompts. The paper reports that this system reaches 53.50% SR and 40.57% SPL on CityAVOS, outperforming indoor-adapted and outdoor baselines by clear margins, while still trailing human operators.

Load-bearing premise

The headline success metric counts an episode as successful when the agent itself stops within 20 meters and says it has found the target, with no independent check that the object is the actual target.

Editorial extensions

If this is right

  • If these results hold, CityAVOS becomes a standard testbed for comparing UAV object-search agents, filling the gap left by indoor-only object navigation and instruction-based aerial navigation benchmarks.
  • PRPSearcher's object-centric semantic segmentation and attraction-based cognitive map suggest that restricting perception to target-related semantics is key to efficient search in cluttered urban scenes.
  • The ablation results imply that neither pure exploration nor pure exploitation is sufficient; the IPT threshold $\theta$ provides a tunable balance that peaks at $\theta=0.1$.
  • The performance gap between the best agent and human operators (53.50% vs 78.68% SR) points to concrete next targets: better semantic reasoning about sparse cues and more efficient spatial exploration.

Reading between the lines

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

  • The paper's success metric relies on the agent's own stop decision without independent verification, so a re-evaluation with a post-hoc object check would likely yield a lower verified success rate; this should be treated as a caveat when comparing headline numbers.
  • Because dataset labels were manually verified without reported inter-annotator agreement, some hard tasks may contain ambiguous or noisy labels; measuring agreement would strengthen the benchmark's reliability.
  • The agent depends on commercial MLLM APIs at every reasoning step, which is costly and slow; distilling the map-building and prompting policy into a smaller model is a natural testable extension.
  • The synthetic Unreal Engine environment may not capture real-world visual noise and drone dynamics, so a sim-to-real transfer study on real drone footage is an open question the paper does not address.
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

3 major / 4 minor

Summary. The paper introduces CityAVOS, a synthetic urban benchmark for aerial visual object search (AVOS) built on the EmbodiedCity simulator, containing 2,420 tasks over six object categories and three difficulty levels. It also proposes PRPSearcher, an MLLM-based agent that constructs an object-centric 3D semantic map, a cognitive map driven by MLLM-assigned attraction values, and a 3D uncertainty map, and that uses an IPT prompting mechanism to balance exploration and exploitation. The authors report that PRPSearcher achieves 53.50% SR and 40.57% SPL on CityAVOS, outperforming several adapted indoor and outdoor baselines, and include ablations over the semantic-map design, the exploration/exploitation components, the IPT threshold, and the choice of MLLM.

Significance. If the reported results hold, the paper would provide a useful standardized benchmark for a relatively underexplored task and a competitive MLLM-agent baseline that is openly described. The construction of a 2,420-task dataset with manual verification, the public code/dataset link, and the ablations are genuine strengths. However, the central performance claims are currently not established: the success criterion is not independently verified, the SPL formula as written is not the standard metric and is not well defined, and the headline numbers are single point estimates with the exploration threshold selected on the same subset used for evaluation. These issues are load-bearing because they directly affect the claimed 53.50% SR and the reported margins over baselines; they are, however, fixable within the scope of the manuscript.

major comments (3)
  1. [Section 3.1, Appendix A.2, Appendix A.3.1] The success flag fs_i is never tied to ground-truth object positions. Section 3.1 states that 'the agent determines whether to search and locate the target successfully based on observations,' Appendix A.2 says the MLLM-based agent decides whether the target has been found and then executes the stop action, and Appendix A.3.1 defines fs_i only as a Boolean without specifying who or what sets it. Appendix A.3.5 mentions 'subsequent verification' in one case study, but no verification protocol is described. Because every counted success therefore rests on the agent's own stop/identification decision, the headline 53.50% SR and the +37.69% margin over baselines may include hallucinated or premature stops. Please add an offline success checker that compares the stop position and detected object against the recorded ground-truth P_object, run it for all methods, and report the corrected metrics.
  2. [Appendix A.3.1, Eq. (19)] The SPL formula SPL = SR * sum_i (tl_i / tl*_i) is not the standard success-weighted path-length metric and is not well defined as written: it multiplies a fraction by a sum of per-episode ratios, so the result can exceed 1 and is not a mean over episodes, and it applies the path ratio to failed episodes as well. This directly affects the reported SPL values of 40.57 for PRPSearcher and all SPL comparisons in Table 2. Please replace it with the standard per-episode definition, typically SPL = (1/q) * sum_i fs_i * (tl*_i / max(tl_i, tl*_i)), clip or otherwise handle tl*_i = 0, and re-report all SPL numbers.
  3. [Section 5.2, Section 5.3, Table 4] All experimental numbers are single point estimates with no variance, repeated runs, or confidence intervals, and the IPT threshold theta_T is tuned on the same 605-task subset that is used for the final evaluation. Table 4 shows a clear performance peak at theta_T = 0.1, but because the same data select and then report this value, the magnitude of the reported improvements and the claim that this setting is 'optimal' are not statistically supported. Please add multiple runs or seeds with standard errors, use a validation split for threshold selection, and state how many episodes each aggregated cell in Tables 2-5 is based on.
minor comments (4)
  1. [Table 2 vs Table 3] The total MSS for PRPSearcher is reported as 35.26 in Table 2 and 35.20 in Table 3; please make the numbers consistent.
  2. [Appendix A.3.1, Eq. (18)] The NE formula uses fp*_i but the surrounding text defines the ground-truth position as tp*; please unify the notation and clarify that tl*_i is the shortest-path length to the target, not the recorded trajectory length.
  3. [Section 4.4 and Table 4] The threshold is written as theta in Eq. (15) but as theta_T in Table 4 and the surrounding text; please use one symbol throughout and define its units or range.
  4. [Section 3.2] The dataset collection is described as manually verified, but no inter-annotator agreement or quality-control statistics are reported; a sentence describing how many trajectories were rejected or regenerated would strengthen confidence in the benchmark labels.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the benchmark numbers are empirical evaluations, not derivations that reduce to their own inputs.

full rationale

The paper's central claims are empirical: CityAVOS is a newly constructed benchmark, and PRPSearcher is evaluated on it. No derivation in the paper equates a predicted quantity to a fitted input by construction. Attraction values in Eq. 6 are produced by an MLLM from task images and text, not fitted to ground-truth target positions, so the cognitive map and exploitation advice are not self-fulfilling. The semantic, cognitive, and uncertainty maps are all constructed from online observations and model inference. Baselines share the same evaluation protocol, so relative comparisons, while possibly affected by metric-design issues, are not circular. The tuning of the IPT threshold θ_T on the evaluation benchmark (Table 4) is a hyperparameter-selection concern rather than a circular step, because selecting a threshold does not force the reported SR to equal the threshold by construction. Self-citations to EmbodiedCity and CityEQA are used as a simulation platform and related prior work, not as an unverified theorem that supplies the paper's conclusions. One genuine gap is that Appendix A.3.1 defines the success flag fs_i only as a Boolean flag and Section 3.1/Appendix A.2 describe the agent itself deciding when the target is found; if no independent oracle or ground-truth distance check sets fs_i, the absolute SR would be a self-reported stop rate. That is an omitted verification specification and a correctness risk, but it is not a circular derivation, because no equation forces the value 53.50% and all compared agents are measured under the same described protocol.

Assumptions & free parameters 5 free parameters · 6 assumptions · 1 invented entities

The paper rests on simulator fidelity, segmentation and MLLM reliability, and an author-defined success metric. It introduces one main invented quantity, the semantic attraction value, which is generated by an MLLM and not independently calibrated. Several hand-chosen parameters affect the reported results, and theta_T is tuned on the same benchmark used for the final evaluation.

free parameters (5)
  • theta_T (exploration advice threshold) = 0.1
    Eq. 15 and Table 4; swept over values on CityAVOS and selected for best SR/SPL, then used in the headline results.
  • alpha (uncertainty attenuation coefficient) = not reported
    Eq. 9; controls uncertainty decay with distance and therefore the exploration reward; no value or sensitivity analysis is given.
  • DBSCAN parameters (eps, min_samples) = not reported
    Section 4.4; clustering of max-attraction grids determines the exploitation target, but the parameters are unspecified.
  • Success radius (20 m) = 20 m
    Section 5.1; hand-chosen threshold defining the success rate, with no sensitivity analysis or external justification.
  • Grid cell size and agent step size = not reported
    Section 4.2 and Appendix A.1; map resolution and action scale are chosen per scene and affect navigation behavior, but exact values are not given.
assumptions (6)
  • standard math Pinhole camera projection with known intrinsics and extrinsics maps depth pixels to world coordinates (Eq. 3).
    Used in Section 4.2 to construct the 3D semantic map; this is standard geometry, not derived in the paper.
  • domain assumption EmbodiedCity/AirSim provides sufficiently realistic RGB-D and pose observations for urban UAV search.
    Section 3.2; all dataset tasks and experiments run in this simulator, with no real-world validation.
  • domain assumption Ground-SAM semantic segmentation produces correct masks and labels for the extracted object semantics.
    Eq. 2; segmentation quality is not evaluated on CityAVOS.
  • domain assumption MLLMs assign reliable attraction values and make reliable stop or identification decisions from images.
    Eq. 6 and the action planning prompt; attraction scores and target identification are MLLM outputs with no calibration or human agreement check.
  • ad hoc to paper Recognized grid cells can be permanently zeroed in the cognitive map via the mirrored map.
    Eq. 8; if a target or clue in a recognized cell was missed, it will not be revisited. This design assumption is not independently validated.
  • ad hoc to paper Stopping within 20 m and identifying the target is a valid success criterion.
    Section 5.1; the threshold and identification requirement are author-defined, and independent verification of identification is not described.
invented entities (1)
  • Semantic attraction value A(s)
    purpose: Quantifies how strongly an observed object's semantics attract the UAV agent toward the target, and drives the exploitation advice.
    Assigned by an MLLM from a prompt (Eq. 6) with no external calibration, ground-truth validation, or human agreement measurement; it is the core of the cognitive map and a central new quantity introduced by the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Autonomous UAV Visual Object Search in City Space: Benchmark and Agentic Methodology." pith.science (2026). https://pith.science/paper/Q6CGFPVR

@misc{pith2026250508765,
  author       = {Pith},
  title        = {Pith review of: Towards Autonomous UAV Visual Object Search in City Space: Benchmark and Agentic Methodology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q6CGFPVR}},
  note         = {Machine review of arXiv:2505.08765}
}
read the original abstract

Aerial Visual Object Search (AVOS) tasks in urban environments require Unmanned Aerial Vehicles (UAVs) to autonomously search for and identify target objects using visual and textual cues without external guidance. Existing approaches struggle in complex urban environments due to redundant semantic processing, similar object distinction, and the exploration-exploitation dilemma. To bridge this gap and support the AVOS task, we introduce CityAVOS, the first benchmark dataset for autonomous search of common urban objects. This dataset comprises 2,420 tasks across six object categories with varying difficulty levels, enabling comprehensive evaluation of UAV agents' search capabilities. To solve the AVOS tasks, we also propose PRPSearcher (Perception-Reasoning-Planning Searcher), a novel agentic method powered by multi-modal large language models (MLLMs) that mimics human three-tier cognition. Specifically, PRPSearcher constructs three specialized maps: an object-centric dynamic semantic map enhancing spatial perception, a 3D cognitive map based on semantic attraction values for target reasoning, and a 3D uncertainty map for balanced exploration-exploitation search. Also, our approach incorporates a denoising mechanism to mitigate interference from similar objects and utilizes an Inspiration Promote Thought (IPT) prompting mechanism for adaptive action planning. Experimental results on CityAVOS demonstrate that PRPSearcher surpasses existing baselines in both success rate and search efficiency (on average: +37.69% SR, +28.96% SPL, -30.69% MSS, and -46.40% NE). While promising, the performance gap compared to humans highlights the need for better semantic reasoning and spatial exploration capabilities in AVOS tasks. This work establishes a foundation for future advances in embodied target search. Dataset and source code are available at https://anonymous.4open.science/r/CityAVOS-3DF8.

Figures

Figures reproduced from arXiv: 2505.08765 by the authors.

Figure 1
Figure 1. An illustration case of a UAV performing the AVOS task in an unfamiliar urban environment. In the search process, [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Examples of six object categories and dataset statistics of the CityAVOS. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Overview of the agentic method–PRPSearcher. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Two selected cases of PRPSearcher on two episodes. One is a successful case, demonstrating how and why our proposed [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 6
Figure 6. Figure 6: Prompt for related semantics. Details on attractions in 3D cognitive map. The 3D cognitive map reflects the attraction of scene semantics to the agent, which essentially stems from the relevance between the semantics and the target object. To obtain the semantics and t…
Figure 7
Figure 7. Figure 7: Prompt for 3D cognitive maps [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]
Figure 8
Figure 8. Figure 8: Prompt for action planning. During the search process, the drone continuously captures RGB images and depth maps from its current pose at each step. The agent first updates a 3D dynamic semantic map using these visual inputs. This involves performing semantic segmentat…
Figure 9
Figure 9. Figure 9: Workflow of the proposed approach—PRPSearcher. [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]
Figure 10
Figure 10. Figure 10: Running process of typical case 1 [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: Running process of typical case 2 [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 25 canonical work pages

  1. [1]

    Peter Anderson, Qi Wu, Damien Teney, Jake Bruce, Mark Johnson, Niko Sünder- hauf, Ian Reid, Stephen Gould, and Anton Van Den Hengel. 2018. Vision-and- language navigation: Interpreting visually-grounded navigation instructions in real environments. In Proceedings of the IEEE conference on computer vision and pattern recognition. 3674–3683

    work page 2018

  2. [2]

    Walid Bousselham, Felix Petersen, Vittorio Ferrari, and Hilde Kuehne. 2024. Grounding everything: Emerging localization properties in vision-language trans- formers. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 3828–3837

    work page 2024

  3. [3]

    Wenzhe Cai, Siyuan Huang, Guangran Cheng, Yuxing Long, Peng Gao, Changyin Sun, and Hao Dong. 2024. Bridging zero-shot object navigation and founda- tion models through pixel-guided navigation skill. In 2024 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 5228–5234

    2024

  4. [4]

    Zhixi Cai, Cristian Rojas Cardenas, Kevin Leo, Chenyuan Zhang, Kal Backman, Hanbing Li, Boying Li, Mahsa Ghorbanali, Stavya Datta, Lizhen Qu, et al. 2024. NEUSIS: A Compositional Neuro-Symbolic Framework for Autonomous Percep- tion, Reasoning, and Planning in Complex UAV Search Missions. arXiv preprint arXiv:2409.10196 (2024)

    arXiv 2024

  5. [5]

    Angel Chang, Angela Dai, Thomas Funkhouser, Maciej Halber, Matthias Niessner, Manolis Savva, Shuran Song, Andy Zeng, and Yinda Zhang. 2017. Matterport3d: Learning from rgb-d data in indoor environments.arXiv preprint arXiv:1709.06158 (2017)

    arXiv 2017

  6. [6]

    Devendra Singh Chaplot, Dhiraj Prakashchand Gandhi, Abhinav Gupta, and Russ R Salakhutdinov. 2020. Object goal navigation using goal-oriented semantic exploration. Advances in Neural Information Processing Systems 33 (2020), 4247– 4258

    2020

  7. [7]

    Shizhe Chen, Thomas Chabal, Ivan Laptev, and Cordelia Schmid. 2023. Object goal navigation with recursive implicit maps. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 7089–7096

    work page 2023

  8. [8]

    Matt Deitke, Eli VanderBilt, Alvaro Herrasti, Luca Weihs, Kiana Ehsani, Jordi Salvador, Winson Han, Eric Kolve, Aniruddha Kembhavi, and Roozbeh Mot- taghi. 2022. ProcTHOR: Large-Scale Embodied AI Using Procedural Generation. Advances in Neural Information Processing Systems 35 (2022), 5982–5994

    2022

Show all 47 references
  1. [9]

    cat-shaped mug

    Vishnu Sashank Dorbala, James F Mullen, and Dinesh Manocha. 2023. Can an embodied agent find your “cat-shaped mug”? llm-based zero-shot object navigation. IEEE Robotics and Automation Letters 9, 5 (2023), 4083–4090

    2023

  2. [10]

    Björn Döschl and Jane Jean Kiam. 2024. Say-REAPEx: An LLM-Modulo UAV Online Planning Framework for Search and Rescue. In 2nd CoRL Workshop on Learning Effective Abstractions for Planning

    2024

  3. [11]

    Heming Du, Lincheng Li, Zi Huang, and Xin Yu. 2023. Object-goal visual naviga- tion via effective exploration of relations among historical navigation states. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion. 2563–2573

    2023

  4. [12]

    Samir Yitzhak Gadre, Mitchell Wortsman, Gabriel Ilharco, Ludwig Schmidt, and Shuran Song. 2023. Cows on pasture: Baselines and benchmarks for language- driven zero-shot object navigation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 23...

    2023

  5. [13]

    Chen Gao, Baining Zhao, Weichen Zhang, Jinzhu Mao, Jun Zhang, Zhiheng Zheng, Fanhang Man, Jianjie Fang, Zile Zhou, Jinqiang Cui, et al. 2024. EmbodiedCity: A Benchmark Platform for Embodied Agent in Real-world City Environment. arXiv preprint arXiv:2410.09604 (2024)

    2024 arXiv

  6. [14]

    Yunpeng Gao, Chenhui Li, Zhongrui You, Junli Liu, Zhen Li, Pengan Chen, Qizhi Chen, Zhonghan Tang, Liansheng Wang, Penghui Yang, et al. 2025. OpenFly: A Versatile Toolchain and Large-scale Benchmark for Aerial Vision-Language Navigation. arXiv preprint arXiv:2502.18041 (2025)

    2025

  7. [15]

    Yunpeng Gao, Zhigang Wang, Linglin Jing, Dong Wang, Xuelong Li, and Bin Zhao. 2024. Aerial Vision-and-Language Navigation via Semantic-Topo-Metric Representation Guided LLM Reasoning. arXiv preprint arXiv:2410.08500 (2024)

    2024 arXiv

  8. [16]

    Yukai Hou, Jin Zhao, Rongqing Zhang, Xiang Cheng, and Liuqing Yang. 2023. UAV swarm cooperative target search: A multi-agent reinforcement learning approach. IEEE Transactions on Intelligent Vehicles 9, 1 (2023), 568–578

    2023

  9. [17]

    Jungdae Lee, Taiki Miyanishi, Shuhei Kurita, Koya Sakamoto, Daichi Azuma, Yutaka Matsuo, and Nakamasa Inoue. 2024. CityNav: Language-Goal Aerial Navigation Dataset with Geographic Information. arXiv preprint arXiv:2406.14240 (2024)

    2024 arXiv

  10. [18]

    Dongfang Liu, Yiming Cui, Zhiwen Cao, and Yingjie Chen. 2020. Indoor naviga- tion for mobile agents: A multimodal vision fusion model. In 2020 international joint conference on neural networks (IJCNN) . IEEE, 1–8

    2020

  11. [19]

    Shubo Liu, Hongsheng Zhang, Yuankai Qi, Peng Wang, Yanning Zhang, and Qi Wu. 2023. Aerialvln: Vision-and-language navigation for uavs. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 15384–15394

    2023

  12. [20]

    Yang Liu, Weixing Chen, Yongjie Bai, Xiaodan Liang, Guanbin Li, Wen Gao, and Liang Lin. 2024. Aligning cyber space with physical world: A comprehensive survey on embodied ai. arXiv preprint arXiv:2407.06886 (2024)

    2024 arXiv

  13. [21]

    Dujun Nie, Xianda Guo, Yiqun Duan, Ruijun Zhang, and Long Chen. 2025. WM- Nav: Integrating Vision-Language Models into World Models for Object Goal Navigation. arXiv preprint arXiv:2503.02247 (2025)

    2025 arXiv

  14. [22]

    Yuankai Qi, Qi Wu, Peter Anderson, Xin Wang, William Yang Wang, Chunhua Shen, and Anton van den Hengel. 2020. Reverie: Remote embodied visual referring expression in real indoor environments. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition ....

    2020

  15. [23]

    Santhosh Kumar Ramakrishnan, Devendra Singh Chaplot, Ziad Al-Halah, Jiten- dra Malik, and Kristen Grauman. 2022. Poni: Potential functions for objectgoal navigation with interaction-free learning. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recog...

    2022

  16. [24]

    Santhosh K Ramakrishnan, Aaron Gokaslan, Erik Wijmans, Oleksandr Maksymets, Alex Clegg, John Turner, Eric Undersander, Wojciech Galuba, An- drew Westbury, Angel X Chang, et al. 2021. Habitat-matterport 3d dataset (hm3d): 1000 large-scale 3d environments for embodied ai.arXiv p...

    2021 arXiv

  17. [25]

    Erich Schubert, Jörg Sander, Martin Ester, Hans Peter Kriegel, and Xiaowei Xu

  18. [26]

    Shital Shah, Debadeepta Dey, Chris Lovett, and Ashish Kapoor. 2018. Airsim: High-fidelity visual and physical simulation for autonomous vehicles. InField and Service Robotics: Results of the 11th International Conference . Springer, 621–635

    2018

  19. [27]

    Ruifeng She and Yanfeng Ouyang. 2021. Efficiency of UAV-based last-mile delivery under congestion in low-altitude air. Transportation Research Part C: Emerging Technologies 122 (2021), 102878

    2021

  20. [28]

    Xiangyu Wang, Donglin Yang, Ziqin Wang, Hohin Kwan, Jinyu Chen, Wenjun Wu, Hongsheng Li, Yue Liao, and Si Liu. 2024. Towards realistic uav vision- language navigation: Platform, benchmark, and methodology. arXiv preprint arXiv:2410.07087 (2024)

    2024 arXiv

  21. [29]

    Erik Wijmans, Abhishek Kadian, Ari Morcos, Stefan Lee, Irfan Essa, Devi Parikh, Manolis Savva, and Dhruv Batra. 2019. Dd-ppo: Learning near-perfect pointgoal navigators from 2.5 billion frames. arXiv preprint arXiv:1911.00357 (2019)

    2019 arXiv

  22. [30]

    Chunxue Wu, Bobo Ju, Yan Wu, Xiao Lin, Naixue Xiong, Guangquan Xu, Hongyan Li, and Xuefeng Liang. 2019. UAV autonomous target search based on deep reinforcement learning in complex disaster scene. IEEe Access 7 (2019), 117227– 117245

    2019

  23. [31]

    Jie Wu, Tianshui Chen, Lishan Huang, Hefeng Wu, Guanbin Li, Ling Tian, and Liang Lin. 2020. Active Object Search. InProceedings of the 28th ACM International Conference on Multimedia. 973–981

    2020

  24. [32]

    Pengying Wu, Yao Mu, Bingxian Wu, Yi Hou, Ji Ma, Shanghang Zhang, and Chang Liu. 2024. Voronav: Voronoi-based zero-shot object navigation with large language model. arXiv preprint arXiv:2401.02695 (2024)

    2024 arXiv

  25. [33]

    Yan Wu, Mingtao Nie, Xiaolei Ma, Yicong Guo, and Xiaoxiong Liu. 2023. Co- evolutionary algorithm-based multi-unmanned aerial vehicle cooperative path planning. Drones 7, 10 (2023), 606

    2023

  26. [34]

    Fei Xia, Amir R Zamir, Zhiyang He, Alexander Sax, Jitendra Malik, and Silvio Savarese. 2018. Gibson env: Real-world perception for embodied agents. In Proceedings of the IEEE conference on computer vision and pattern recognition . 9068–9079

    2018

  27. [35]

    Linjie Xing, Xiaoyan Fan, Yaxin Dong, Zenghui Xiong, Lin Xing, Yang Yang, Haicheng Bai, and Chengjiang Zhou. 2022. Multi-UAV cooperative system for search and rescue based on YOLOv5. International Journal of Disaster Risk Reduction 76 (2022), 102972

    2022

  28. [36]

    Karmesh Yadav, Ram Ramrakhya, Santhosh Kumar Ramakrishnan, Theo Gervet, John Turner, Aaron Gokaslan, Noah Maestre, Angel Xuan Chang, Dhruv Batra, Manolis Savva, et al. 2023. Habitat-matterport 3d semantics dataset. InProceedings of the IEEE/CVF Conference on Computer Vision an...

    2023

  29. [37]

    Fanglong Yao, Yuanchang Yue, Youzhi Liu, Xian Sun, and Kun Fu. 2024. Aero- verse: Uav-agent benchmark suite for simulating, pre-training, finetuning, and evaluating aerospace embodied world models. arXiv preprint arXiv:2408.15511 (2024)

    2024

  30. [38]

    Joel Ye, Dhruv Batra, Abhishek Das, and Erik Wijmans. 2021. Auxiliary tasks and exploration enable objectgoal navigation. In Proceedings of the IEEE/CVF international conference on computer vision . 16117–16126

    2021

  31. [39]

    Bangguo Yu, Hamidreza Kasaei, and Ming Cao. 2023. Co-navgpt: Multi-robot cooperative visual semantic navigation using large language models. arXiv preprint arXiv:2310.07937 (2023)

    2023 arXiv

  32. [40]

    Bangguo Yu, Hamidreza Kasaei, and Ming Cao. 2023. L3mvn: Leveraging large language models for visual target navigation. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 3554–3560

    2023

  33. [41]

    Jie Zhang, Mingxuan Li, Yitai Xu, Hua He, Qun Li, and Tao Wang. 2025. StrucGCN: Structural enhanced graph convolutional networks for graph embedding. Infor- mation Fusion 117 (2025), 102893

    2025

  34. [42]

    Nan Zhao, Weidang Lu, Min Sheng, Yunfei Chen, Jie Tang, F Richard Yu, and Kai-Kit Wong. 2019. UAV-assisted emergency networks in disasters.IEEE Wireless Communications 26, 1 (2019), 45–51

    2019

  35. [43]

    Yong Zhao, Bin Chen, XiangHan Wang, Zhengqiu Zhu, Yiduo Wang, Guangquan Cheng, Rui Wang, Rongxiao Wang, Ming He, and Yu Liu. 2022. A deep rein- forcement learning based searching method for source localization. Information Sciences 588 (2022), 67–81

    2022

  36. [44]

    Yong Zhao, Kai Xu, Zhengqiu Zhu, Yue Hu, Zhiheng Zheng, Yingfeng Chen, Yatai Ji, Chen Gao, Yong Li, and Jincai Huang. 2025. Cityeqa: A hierarchical llm agent on embodied question answering benchmark in city space. arXiv preprint arXiv:2502.12532 (2025)

    2025 arXiv

  37. [45]

    Gengze Zhou, Yicong Hong, and Qi Wu. 2024. Navgpt: Explicit reasoning in vision-and-language navigation with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 38. 7641–7649

    2024

  38. [46]

    Kaiwen Zhou, Kaizhi Zheng, Connor Pryor, Yilin Shen, Hongxia Jin, Lise Getoor, and Xin Eric Wang. 2023. Esc: Exploration with soft commonsense constraints for zero-shot object navigation. In International Conference on Machine Learning . PMLR, 42829–42842. Conference’17, July ...

    2023

  39. [2017]

    ACM Transactions on Database Systems (TODS) 42, 3 (2017), 1–21

    DBSCAN revisited, revisited: why and how you should (still) use DBSCAN. ACM Transactions on Database Systems (TODS) 42, 3 (2017), 1–21. Towards Autonomous UAV Visual Object Search in City Space: Benchmark and Agentic Methodology Conference’17, July 2017, Washington, DC, USA

    2017

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.