Pith. sign in

REVIEW 6 major objections 8 minor 26 references

Tactical Decision for Multi-UGV Confrontation with a Vision-Language Model-Based Commander

T0 review · 6 major / 8 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A vision-language model commander that separates perception from planning beats rule-based and reinforcement-learning baselines in simulated UGV confrontations.

desk verdict A sensible VLM+LLM commander that beats its own rule-based teacher in a bespoke simulator, but the expert system doubles as the baseline and the thresholds are hidden, so the headline win rate is not yet established. read the letter →

arxiv 2507.11079 v1 pith:QUMH2KZB submitted 2025-07-15 cs.AI

Li Wang , Qizhen Wu , Lei Chen This is my paper · ORCID
classification cs.AI
keywords multi-agentconfrontationvision-languagemodellargelanguagetacticaldecisionunmannedgroundvehiclesexpertsysteminterpretableAIreinforcementlearningbaseline
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

The paper tries to establish that a commander built from a vision-language model (VLM) for scene understanding and a lightweight large language model (LLM) for tactical planning can make interpretable, adaptive decisions in multi-UGV confrontations. It proposes a full-chain pipeline: the VLM converts a bird's-eye image into hierarchical semantic descriptions, and the LLM outputs per-agent instructions such as attack, support, or retreat. An expert system, active only during training, generates rule-based labels and preference pairs that align the two modules. In a 30 m by 16 m simulator with 5-on-5 UGVs, the approach reports win rates of 83 percent against a rule-based enemy and 80 percent against a reinforcement-learning enemy, with higher survival rates and lower decision times than a single-VLM baseline. If correct, the result supports the idea that separating perception from strategic reasoning, with shared semantic language, is a practical route to transparent tactical AI.

What carries the argument

The load-bearing mechanism is the expert system: a set of heuristic rules (Table 1) built on three computed metrics—threat score $T(e_j)$ from Eq. 4, danger value $D(a_i)$ from Eq. 6, and attack cost $C(a_i,e_j)$ from Eqs. 7–8—that convert structured perception into tactical labels. During training, these labels supervise the VLM's semantic outputs and provide chain-of-thought exemplars plus DPO preference pairs for the LLM, aligning the two modules within a shared semantic space $S = \{S_u, S_l, S_r\}$ at unit, local, and region levels. The same metric-driven rules also serve as the Rule baseline in the comparison, which is why the reported win rates should be read as a comparison against the authors' own rule set.

What would settle it

Run the same VLM-plus-LLM commander against an expert system whose thresholds and weights are tuned adversarially, or against a human expert, and measure whether the reported 80 percent-plus win rate survives; a drop to near 50 percent would indicate the advantage comes from the specific rule set rather than from the learned reasoning.

Watch

Extended reading notes

Core claim

The central claim is that a VLM-plus-LLM commander, trained with an expert system's rule-based supervision, outperforms both handcrafted rules and a single VLM in autonomous ground-vehicle confrontations. The paper reports, in Table 2, an 83 percent win rate against the Rule enemy and 80 percent against the RL enemy, a survival rate of 0.90 and 0.86 respectively, and an average decision time of about 15.4 seconds and 14.5 seconds, roughly 25 percent faster than the single-VLM baseline. The authors interpret this as evidence that offloading tactical reasoning to a language planner lets the visual module focus on perception, improving both accuracy and speed, and that the LLM can tolerate perception noise and produce strategic formations such as encirclement. They also report that ablations removing the LLM or the expert system degrade performance, and that the advantage persists when the scenario is scaled from 5-on-5 to 7-on-7 and 9-on-9.

Load-bearing premise

The load-bearing premise is that the expert system's handcrafted rules encode genuinely good tactics, because those same rules generate the training labels and also serve as the Rule baseline that the method is compared against.

Editorial extensions

If this is right

  • If the reported win rates hold, separating VLM perception from LLM planning is a viable architecture for adversarial multi-agent command, offering a balance between computational cost and decision quality.
  • The modular design means perception and planning can be upgraded independently: a better VLM or a stronger LLM should improve performance without changing the semantic interface.
  • The expert-system training paradigm offers a way to inject human tactical knowledge into LLM decision-making without hand-designing every rule, since the LLM can generalize beyond the rule set.
  • The observed tolerance to perception noise suggests that robust tactical reasoning can partially compensate for imperfect localization, which matters for real deployments where perception errors are inevitable.
  • Scaling results from 5-on-5 to 7-on-7 and 9-on-9 indicate the approach maintains its advantage as agent density grows, although decision time increases with scale.

Reading between the lines

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

  • The win criterion in the simulator is complete elimination of one team; real engagements often involve partial objectives, asymmetric capabilities, or retreat rules, so the 80 percent figure does not directly transfer to deployment settings.
  • Because the expert system is also the Rule baseline, the comparison measures the gap between a learned policy and its own teacher; a stronger independent baseline (for example, a different rule set or a human expert) would clarify how much genuine tactical novelty the LLM adds.
  • The semantic interface could be ported to other multi-agent adversarial domains where bird's-eye perception is available, but the expert system's equations and thresholds would need re-derivation for those dynamics, so the method is not immediately drop-in general.
  • A testable extension is replacing the expert system with a learned reward model or on-policy reinforcement learning fine-tuning, which would test whether rule-based supervision is necessary or merely convenient.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

6 major / 8 minor

Summary. The paper proposes a hierarchical commander for multi-UGV confrontation, combining a vision-language model (VLM) for scene understanding and a lightweight large language model (LLM) for tactical reasoning. The two modules communicate through a structured semantic space, and the system is trained using labels generated by a handcrafted expert rule system (Section 3.3). Simulation experiments with 5v5 confrontations report win rates of 83% against a rule-based enemy and 80% against a reinforcement-learning enemy, with additional ablations and a generalization study to 7v7 and 9v9 settings. The authors argue that separating perception from reasoning improves perceptual accuracy, reduces decision time, and yields better tactical performance than single-VLM or rule-based baselines.

Significance. If the reported results are robust, the paper demonstrates a plausible and interpretable architecture for tactical decision-making in adversarial multi-agent scenarios, with a clean separation of perception and reasoning via a shared semantic space. The use of a lightweight LLM as a strategic planner is timely and the modular design is appealing. However, the current evidence is not yet convincing: the central comparison is potentially circular because the expert rule system that generates training labels also serves as the Rule baseline; statistical uncertainty is not reported despite only 10 episodes per condition; the RL baseline is not described; and the generalization results are unreadable in the provided figure. These issues prevent the paper from establishing its headline 'over 80% win rate' claim as a general result. The paper does provide a concrete experimental pipeline and an honest discussion of perception limitations, which are strengths, but the validation needs substantial strengthening.

major comments (6)
  1. [Section 3.3, Table 1, and Section 4.2, Table 2] The expert system described in Section 3.3 (Eqs. 4–8 and Table 1) generates the training labels for the proposed commander and also serves as the 'Rule' baseline in Table 2. This creates a circular comparison: the learned model is trained to imitate these very rules, so a higher win rate against the Rule baseline chiefly indicates that the model has learned to deviate from its teacher, not that the tactics are strong in any absolute sense. The paper should provide an independent tactical reference (e.g., a standard baseline from the literature, a human-expert evaluation, or a much stronger rule set) and a sensitivity analysis of the expert-system parameters to demonstrate that the reported advantage is not an artifact of a weak or idiosyncratic rule implementation.
  2. [Section 4.1 and Table 2] The experimental setup states that each condition is repeated with ten independent experiments, but Table 2 reports only point estimates of win rate, survival rate, and decision time, with no variance, confidence intervals, or significance tests. With n=10 episodes per condition, the differences between, for example, 83% and 67% win rates (vs. Rule) or 80% and 71% (vs. RL) may be within sampling noise. The authors should report standard deviations or confidence intervals and perform a suitable statistical test (e.g., bootstrap or Mann–Whitney U) to support the claim of superiority.
  3. [Section 4.2, Table 2] The reinforcement-learning baseline (denoted 'RL') is never described. The paper does not state which RL algorithm is used, the network architecture, the observation and action spaces, the reward function, or any training hyperparameters. Without this information, the comparison 'Ours vs. RL' is uninterpretable, and the reader cannot assess whether the RL baseline is a strong or weak representative of the method class. The authors must provide a complete description of the RL baseline, consistent with how the VLM baseline is described.
  4. [Section 4.2, Table 2] Perception metrics P, R, and RH (perception accuracy, recall rate, and hallucination rate) are mentioned but never formally defined. In particular, it is unclear what constitutes a 'hallucination' in this context, how ground truth is obtained for the perception evaluation, and over which units or frames the precision and recall are computed. Without precise definitions, these metrics cannot be reproduced or meaningfully compared across systems.
  5. [Figure 6 and Section 4.3] Figure 6, which reports the generalization results for 7v7 and 9v9 scenarios, is unreadable: the figure text is corrupted in the manuscript, and the underlying numerical values are not given in the text or a table. Since the robustness of the method at larger scales is one of the paper's claims, the authors must provide a clean, legible figure and report the exact numbers for perception accuracy, decision time, survival rate, and decision gain at each scale.
  6. [Equation (4)] The threat score T(e_j) is defined as the total Gaussian threat field I evaluated at the position of e_j, where I sums contributions from all enemies el (including e_j itself). This makes T(e_j) a measure of the ambient threat at that location rather than the strategic impact attributable to enemy e_j. Because T(e_j) is subsequently used in the attack-cost rules (e.g., R8 in Table 1), the definition should be corrected to isolate e_j's own contribution, or the authors should explicitly explain why the aggregate field is the intended interpretation.
minor comments (8)
  1. [Abstract and Table 2] The abstract claims a 'win rate of over 80%', but Table 2 reports exactly 80% against the RL enemy; consider using 'at least 80%' or 'over 80% in one scenario and 80% in the other' to match the data.
  2. [Section 3.3, Equations (4)–(8)] Many thresholds and weights (θ, δ, ε, ω, γ, ζ, σ) are introduced without numerical values or a sensitivity analysis. Providing the chosen values (or a repository with configuration files) would greatly improve reproducibility.
  3. [Equations (4) and (5)] The symbol σ is used both as the Gaussian bandwidth in Eq. (4) and as the logistic function in Eq. (5); the dual use is confusing and should be disambiguated.
  4. [Section 4.1] Equation (4) uses an aspect-ratio parameter ρ for the map, but the arena dimensions are given (30 m × 16 m) and ρ is not defined; please clarify its value or relation to the arena.
  5. [Section 3.4] The fine-tuning description mentions LoRA adaptation, DPO alignment, and cosine learning rate schedules, but omits concrete hyperparameters (e.g., LoRA rank, learning rate, batch size, number of epochs). These details are necessary for reproducibility.
  6. [Table 2] The rows for the Rule-based allied model (vs. Rule and vs. RL) show dashes for all metrics; the authors should indicate that these are not applicable because the rule model has no perception module, or provide equivalent numbers where possible.
  7. [Figures 4 and 5] The figures that illustrate a confrontation example and the survival curves are difficult to read in the current resolution; enlarging the panels or adding higher-resolution versions would help the reader verify the qualitative claims.
  8. [General] There are several minor grammatical and stylistic issues (e.g., 'the' vs. 'a', missing spaces) and some reference entries lack page ranges; a careful copyedit would improve readability.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the expert-system-as-baseline overlap is a baseline-validity limitation, not a circular derivation.

full rationale

The paper's central claim is an empirical win-rate comparison, not a derivation from the expert system. The expert system (Section 3.3, Eqs. 4-8, Table 1) generates supervised training labels, and the same rule set is used as the 'Rule' baseline (Section 4.2, Table 2). This creates a baseline-independence limitation: the 83% win rate versus Rule measures how well the learned commander outperforms its own teacher, and the unstated thresholds and lack of sensitivity analysis weaken external validity. However, this is not circular in the technical sense: no equation defines the reported win rate in terms of the training labels, the model is not forced to match the rules, and the same table reports an 80% win rate against an RL enemy while ablations (Table 3) show the full model beats a single VLM trained with the same expert (67%->83% vs Rule; 71%->80% vs RL). The only self-citation, reference [5] by two of the authors, appears in a related-work sentence about hierarchical RL limitations and is not load-bearing. Therefore no circular step can be exhibited under the stated rules.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claim rests on the author-defined expert system, whose weights and thresholds are undisclosed free parameters. The simulation environment and semantic representation are domain assumptions. No new physical entities are introduced, but the expert system acts as an invented component with no independent evidence.

free parameters (4)
  • Expert system weights: wen, wal, wd,en, wd,al = unspecified
    Equation (6) tunes enemy/ally counts and distance effects in danger values; no numeric values are given.
  • Expert system weights: ωd, ωo, ωt, ωvis, γ, ζ = unspecified
    Equations (7)-(8) define attack and cooperative engagement costs; all weights are stated as positive tuning parameters but never assigned.
  • Expert system thresholds: θcost, θdist, θsep, θcoop, δattack, δthreat, δcontain, δcost, ϵengage, ϵthreat = unspecified
    Table 1's rules depend on these thresholds; without values the expert system's behavior is underdetermined.
  • Gaussian bandwidth σ and radius r in threat/danger fields = unspecified
    Equations (4)-(5) use a Gaussian field and local force balance radius; σ and r are not quantified.
assumptions (4)
  • domain assumption The simulator's physics and engagement rules are representative of real multi-UGV confrontation.
    Section 4.1 defines a 30m x 16m arena, 1m attack radius, 0.3m obstacles, 2 m/s speed cap; no validation that results transfer to real or other simulated conditions.
  • ad hoc to paper The expert system's heuristic rules generate high-quality tactical decisions that are a suitable supervision signal.
    Section 3.3 states the expert system 'acts as an agent for human reasoning' without comparing it to any ground-truth optimal policy or external benchmark.
  • ad hoc to paper The hierarchical semantic output (unit, local, region) captures all information needed for tactical planning.
    Section 3.2 defines S={Su,Sl,Sr}; the completeness of this representation is asserted, not proven.
  • domain assumption QWEN2.5-VL-7B and QWEN2.5-3B provide adequate pretrained knowledge for this task.
    Section 3.4 selects these backbones; no comparison to other model choices or discussion of whether their pretraining is suited to tactical imagery.
invented entities (1)
  • Expert system decision module
    purpose: Generates high-quality decision labels for VLM-LLM training and serves as the Rule baseline in evaluation.
    The expert system is an author-defined rule set with no external validation or established precedent; its quality as a supervisor is assumed, and its weights are undisclosed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tactical Decision for Multi-UGV Confrontation with a Vision-Language Model-Based Commander." pith.science (2026). https://pith.science/paper/QUMH2KZB

@misc{pith2026250711079,
  author       = {Pith},
  title        = {Pith review of: Tactical Decision for Multi-UGV Confrontation with a Vision-Language Model-Based Commander},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QUMH2KZB}},
  note         = {Machine review of arXiv:2507.11079}
}
read the original abstract

In multiple unmanned ground vehicle confrontations, autonomously evolving multi-agent tactical decisions from situational awareness remain a significant challenge. Traditional handcraft rule-based methods become vulnerable in the complicated and transient battlefield environment, and current reinforcement learning methods mainly focus on action manipulation instead of strategic decisions due to lack of interpretability. Here, we propose a vision-language model-based commander to address the issue of intelligent perception-to-decision reasoning in autonomous confrontations. Our method integrates a vision language model for scene understanding and a lightweight large language model for strategic reasoning, achieving unified perception and decision within a shared semantic space, with strong adaptability and interpretability. Unlike rule-based search and reinforcement learning methods, the combination of the two modules establishes a full-chain process, reflecting the cognitive process of human commanders. Simulation and ablation experiments validate that the proposed approach achieves a win rate of over 80% compared with baseline models.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 12 canonical work pages

  1. [1]

    IEEE Trans

    Hou, Y., Liang, X., Zhang, J., Lv, M., Yang, A.: Hierarchical decision-making framework for multiple ucavs autonomous confrontation. IEEE Trans. Veh. Tech- nol. 72(11), 13953–13968 (2023) https://doi.org/10.1109/TVT.2023.3285223

    arXiv 2023

  2. [2]

    IEEE Trans

    Liu, D., Zong, Q., Zhang, X., Zhang, R., Dou, L., Tian, B.: Game of drones: Intel- ligent online decision making of multi-uav confrontation. IEEE Trans. Emerg. Top. Comput. Intell. 8(2), 2086–2100 (2024) https://doi.org/10.1109/TETCI. 2024.3360282

    arXiv 2024

  3. [3]

    Haarnoja, T., Moran, B., Lever, G., Huang, S.H., Tirumala, D., Humplik, J., Wulfmeier, M., Tunyasuvunakool, S., Siegel, N.Y., Hafner, R., et al.: Learning agile soccer skills for a bipedal robot with deep reinforcement learning. Sci. Robot. 9(89), 8022 (2024) https://doi.org/10.1126/scirobotics.adi8022

    doi:10.1126/scirobotics.adi8022 2024

  4. [4]

    Ocean Eng

    Qu, X., Gan, W., Song, D., Zhou, L.: Pursuit-evasion game strategy of usv based on deep reinforcement learning in complex multi-obstacle environment. Ocean Eng. 273, 114016 (2023) https://doi.org/10.1016/j.oceaneng.2023.114016

    arXiv 2023

  5. [5]

    IEEE Trans

    Wu, Q., Liu, K., Chen, L., L¨ u, J.: Hierarchical reinforcement learning for swarm confrontation with high uncertainty. IEEE Trans. Autom. Sci. Eng.22, 8630–8644 (2024) https://doi.org/10.1109/TASE.2024.3487219

    arXiv 2024

  6. [6]

    Nian, X., Li, M., Wang, H., Gong, Y., Xiong, H.: Large-scale uav swarm con- frontation based on hierarchical attention actor-critic algorithm. Appl. Intell. 54, 3279–3294 (2024) https://doi.org/10.1007/s10489-024-05293-5

    work page doi:10.1007/s10489-024-05293-5 2024

  7. [7]

    IEEE Robot

    Mei, A., Zhu, G.-N., Zhang, H., Gan, Z.: Replanvlm: Replanning robotic tasks with visual language models. IEEE Robot. Autom. Lett. 9(11), 10201–10208 (2024) https://doi.org/10.1109/LRA.2024.3471457

    work page arXiv 2024

  8. [8]

    Robot World Cup15570, 176–188 (2025) https://doi.org/10.1007/ 978-3-031-85859-8 15

    Brienza, M., Musumeci, E., Suriani, V., Affinita, D., Pennisi, A., Nardi, D., Bloisi, D.D.: Llcoach: Generating robot soccer plans using multi-role large lan- guage models. Robot World Cup15570, 176–188 (2025) https://doi.org/10.1007/ 978-3-031-85859-8 15

    work page 2025

Show all 26 references
  1. [9]

    Tian, Y., Lin, F., Li, Y., Zhang, T., Zhang, Q., Fu, X., Huang, J., Dai, X., Wang, Y., Tian, C., et al.: Uavs meet llms: Overviews and perspectives towards agentic low-altitude mobility. Inf. Fusion 122, 103158 (2025) https://doi.org/10.1016/j. inffus.2025.103158 15

    2025

  2. [10]

    Science 378(6624), 1067–1074 (2022) https://doi.org/10.1126/science.ade9097

    (F AIR), M.F.A.R.D.T., Bakhtin, A., Brown, N., Dinan, E., Farina, G., Flaherty, C., Fried, D., Goff, A., Gray, J., Hu, H., et al.: Human-level play in the game of diplomacy by combining language models with strategic reasoning. Science 378(6624), 1067–1074 (2022) https://doi.o...

    2022 doi

  3. [11]

    Fontana, N., Pierri, F., Aiello, L.M.: Nicer than humans: How do large language models behave in the prisoner’s dilemma? Proc. Int. AAAI Conf. Web Soc. Media 19(1), 522–535 (2025) https://doi.org/10.1609/icwsm.v19i1.35829

    2025 doi

  4. [12]

    Jin, X., Wang, Z., Du, Y., Fang, M., Zhang, H., Wang, J.: Learning to dis- cuss strategically: A case study on one night ultimate werewolf. Adv. Neural Inf. Process. Syst. 37, 77060–77097 (2024)

    2024

  5. [13]

    CoRR abs/2310.01320 (2023) https://doi.org/10.48550/ arXiv.2310.01320

    Wang, S., Liu, C., Zheng, Z., Qi, S., Chen, S., Yang, Q., Zhao, A., Wang, C., Song, S., Huang, G.: Avalon’s game of thoughts: Battle against deception through recur- sive contemplation. CoRR abs/2310.01320 (2023) https://doi.org/10.48550/ arXiv.2310.01320

    2023 doi arXiv

  6. [14]

    J Intell Robot Syst 109(4), 91 (2023) https: //doi.org/10.1007/s10846-023-01991-3

    Bottega, J.A., Kich, V.A., Jesus, J.C.d., Steinmetz, R., Kolling, A.H., Grando, R.B., Guerra, R.d.S., Gamarra, D.F.T.: Jubileo: an immersive simulation frame- work for social robot design. J Intell Robot Syst 109(4), 91 (2023) https: //doi.org/10.1007/s10846-023-01991-3

    2023 doi

  7. [15]

    Zhai, Y., Bai, H., Lin, Z., Pan, J., Tong, S., Zhou, Y., Suhr, A., Xie, S., LeCun, Y., Ma, Y., Levine, S.: Fine-tuning large vision-language models as decision-making agents via reinforcement learning. Proc. NeurIPS 37, 110935–110971 (2024)

    2024

  8. [16]

    Mei, A., Wang, J., Zhu, G.-N., Gan, Z.: Gamevlm: A decision-making framework for robotic task planning based on visual language models and zero-sum games. Proc. IEEE Int. Conf. Mechatronics Autom., 1771–1776 (2024) https://doi.org/ 10.1109/ICMA61710.2024.10633088

    2024

  9. [17]

    ACM Comput

    Zhang, Z., Yao, Y., Zhang, A., Tang, X., Ma, X., He, Z., Wang, Y., Gerstein, M., Wang, R., Liu, G., Zhao, H.: Igniting language intelligence: The hitchhiker’s guide from chain-of-thought reasoning to language agents. ACM Comput. Surv. 57(8), 39 (2025) https://doi.org/10.1145/3719341

    2025 doi

  10. [18]

    Mu, Y., Zhang, Q., Hu, M., Wang, W., Ding, M., Jin, J., Wang, B., Dai, J., Qiao, Y., Luo, P.: Embodiedgpt: Vision-language pre-training via embodied chain of thought. Adv. Neural Inf. Process. Syst. 36, 25081–25094 (2023)

    2023

  11. [19]

    J Intell Robot Syst 109(4), 76 (2023) https://doi.org/10.1007/ s10846-023-02011-0

    Skaltsis, G.M., Shin, H.-S., Tsourdos, A.: A review of task allocation meth- ods for uavs. J Intell Robot Syst 109(4), 76 (2023) https://doi.org/10.1007/ s10846-023-02011-0

    2023

  12. [20]

    Neurocomputing 580, 127505 (2024) 16 https://doi.org/10.1016/j.neucom.2024.127505

    Liu, C., Wang, C., Peng, Y., Li, Z.: Zvqaf: Zero-shot visual question answering with feedback from large language models. Neurocomputing 580, 127505 (2024) 16 https://doi.org/10.1016/j.neucom.2024.127505

    2024

  13. [21]

    ACM Trans

    Jiang, X., Dong, Y., Wang, L., Fang, Z., Shang, Q., Li, G., Jin, Z., Jiao, W.: Self- planning code generation with large language models. ACM Trans. Softw. Eng. Methodol. 33(7), 1–30 (2024) https://doi.org/10.1145/3672456

    2024 doi

  14. [22]

    Kannan, S.S., Venkatesh, V.L.N., Min, B.-C.: Smart-llm: Smart multi-agent robot task planning using large language models. Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst., 12140–12147 (2024) https://doi.org/10.1109/IROS58592. 2024.10802322

    2024

  15. [23]

    J Intell Robot Syst 110(3), 123 (2024) https://doi.org/10.1007/s10846-024-02159-3

    Zager, M., Sieber, C., Fay, A.: Towards semantic interoperability: An information model for autonomous mobile robots. J Intell Robot Syst 110(3), 123 (2024) https://doi.org/10.1007/s10846-024-02159-3

    2024 doi

  16. [24]

    IEEE Robot

    Song, D., Liang, J., Payandeh, A., Raj, A.H., Xiao, X., Manocha, D.: Vlm- social-nav: Socially aware robot navigation through scoring using vision-language models. IEEE Robot. Autom. Lett. 10(1), 508–515 (2025) https://doi.org/10. 1109/LRA.2024.3511409

    2025

  17. [25]

    cat- shaped mug

    Dorbala, V.S., Mullen, J.F., Manocha, D.: Can an embodied agent find your “cat- shaped mug”? llm-based zero-shot object navigation. IEEE Robot. Autom. Lett. 9(5), 4083–4090 (2024) https://doi.org/10.1109/LRA.2023.3346800

    2024

  18. [26]

    J Intell Robot Syst 107(4), 56 (2023) https: //doi.org/10.1007/s10846-023-01856-9 17

    Yin, C., Zhang, Q.: A multi-modal framework for robots to learn manipulation tasks from human demonstrations. J Intell Robot Syst 107(4), 56 (2023) https: //doi.org/10.1007/s10846-023-01856-9 17

    2023 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.