The reviewed record of science sign in
Pith

arxiv: 2605.25423 · v1 · pith:LEMIQU6W · submitted 2026-05-25 · cs.RO

OPAL: Omnidirectional Path-efficient Aerial 3D expLoration

Yoga Satwik Chappidi , Avideh Zakhor This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-06-29 21:48 UTCgrok-4.3pith:LEMIQU6Wrecord.jsonopen to challenge →

classification cs.RO
keywords aerial robot exploration3D mappingpath planningfrontier selectiondrone navigationomnidirectional yawexploration efficiency
0
0 comments X

The pith

OPAL uses deliberate 360-degree yaw rotations at branch points to achieve shorter travel distances in aerial 3D exploration than global planning methods.

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

The paper introduces OPAL, an exploration method for aerial robots that performs deliberate 360-degree yaw rotations at points where the best direction is unclear. This gathers omnidirectional information to select better frontiers for mapping, avoiding the computational cost of global tour planning used in other approaches. Simulations against EDEN and FALCON show OPAL travels shorter distances with higher coverage per distance, though total time is longer due to rotations. Real-world tests on a drone in indoor spaces confirm a variant travels as much as 25% less distance than FALCON. Variants differ in how they choose frontiers after the rotation, including using LLMs or VLMs, and the search radius can be tuned for distance versus time tradeoffs.

Core claim

By centering exploration on in-place 360-degree yaw rotations at ambiguous branch points rather than compute-heavy global tour planning, OPAL variants achieve shorter travel distances and higher coverage-versus-distance area under the curve in simulations, with real drone experiments showing up to 25% lower traveled distance than FALCON.

What carries the argument

Deliberate 360-degree yaw rotation at ambiguous branch points to obtain omnidirectional views for improved local frontier selection.

If this is right

  • Adjusting the vicinity search radius allows trading off travel distance against total exploration time.
  • OPAL is computationally simpler than baselines like EDEN and FALCON while maintaining shorter paths.
  • Model-free, LLM-based, and VLM-based frontier selection all leverage the yaw rotation for better performance.
  • Real-world validation on a Modal AI drone in two indoor environments confirms the distance reductions.

Where Pith is reading between the lines

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

  • This local rotation strategy may suit resource-limited drones where global optimization exceeds onboard compute.
  • The approach could extend to other platforms where rotation costs less than translation.
  • Testing in larger or outdoor spaces would reveal whether branch-point rotations still yield net distance savings.

Load-bearing premise

That the extra time for 360-degree yaw rotations is offset by sufficiently improved frontier choices that reduce overall travel distance compared to more complex planners.

What would settle it

A test in the same environments showing that OPAL does not achieve lower traveled distance than FALCON or higher coverage per distance despite the added rotation time.

Figures

Figures reproduced from arXiv: 2605.25423 by Avideh Zakhor, Yoga Satwik Chappidi.

Figure 1
Figure 1. Figure 1: Overview of OPAL with exploration stack inherited from FALCON. The orange group is the onboard sensing layer, the cyan group is the base [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Example OPAL decision point map slab after reachable-vicinity [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: OPAL-V prompt template and sample JSON response. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Map 666 from the HM3D benchmark. (a) The raw HM3D scene [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: OPAL selector and geometric-policy comparison. (a) Coverage [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Reachable-vicinity radius rv sweep across the OPAL-NFP family and SoTA baselines. (a) Coverage-distance AUC, where higher is better. (b) Total traversed distance to full coverage, where lower is better, with reference values ranging from 115 m to 266 m. (c) Average elapsed duration by map, where lower is better, with reference values ranging from 112 s to 204 s. (d) Elapsed duration after removing 360◦ pan… view at source ↗
Figure 7
Figure 7. Figure 7: Hardware-validation setup. (a) The ModalAI Starling 2 platform [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Representative voxel maps from hardware runs in Indoor Layout 2. [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
read the original abstract

Autonomous exploration is critical for robot mapping unknown environments. Desirable characteristics of exploration algorithms include compute efficiency and small traversed distance during the exploration process. Motivated by these, we present Omnidirectional Path-efficient Aerial 3D expLoration (OPAL), an exploration framework centered on deliberate 360-degree yaw rotation at ambiguous branch points rather than compute-heavy global tour planning. We devise multiple variants of OPAL to determine the frontier-selection strategy once the yaw pan is completed. One variant is model-free, while others use large language models (LLMs) or vision-language models (VLMs). We characterize the performance of these variants while varying the vicinity search radius to include frontiers in the selection process. Through simulations we find that although the time-consuming in-place yaw rotation increases total exploration time relative to more computationally complex baselines such as EDEN and FALCON, OPAL is computationally simpler and achieves shorter travel distances and higher coverage-versus-distance area under the curve. We also show that adjusting the frontier-selection search radius enables a tradeoff between travel distance and total exploration time. We verify our results on a Modal AI drone in two indoor environments by comparing OPAL against FALCON, and find that the traveled distance for a variant of OPAL to be as much as 25% lower than FALCON.

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

1 major / 1 minor

Summary. The paper introduces OPAL, an aerial 3D exploration framework that performs deliberate 360° in-place yaw rotations at ambiguous branch points to gather additional information for frontier selection, rather than relying on compute-intensive global tour planning. Multiple variants are presented (model-free, LLM-based, VLM-based) with a tunable vicinity search radius; simulations compare against EDEN and FALCON on distance, coverage-vs-distance AUC, and total time, while real-world tests on a Modal AI drone in two indoor environments report up to 25% lower traveled distance for one OPAL variant versus FALCON.

Significance. If the empirical claims hold, OPAL demonstrates that a lightweight local strategy with occasional rotations can achieve shorter paths than global planners while remaining computationally simpler, which is valuable for resource-constrained aerial robots. The simulation results on distance and AUC metrics, plus the real-world distance comparison, provide concrete evidence of the tradeoff; the ability to tune the search radius for distance-vs-time balance is a practical contribution.

major comments (1)
  1. [Abstract / real-world experiments] Abstract (hardware verification paragraph) and real-world experiments section: only traveled distance (up to 25% lower than FALCON) is reported for the Modal AI drone trials in two indoor environments; total exploration time, coverage completion time, or yaw-rotation overhead are omitted. This is load-bearing for the central claim because the abstract and simulation results explicitly note that yaw rotations increase total time relative to baselines, yet the practical net benefit (distance savings outweighing rotation duration) cannot be assessed without the corresponding time metrics in hardware.
minor comments (1)
  1. [Abstract] Abstract: the sentence 'find that the traveled distance for a variant of OPAL to be as much as 25% lower than FALCON' is grammatically incomplete (missing 'is').

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on the real-world experiments. The point regarding missing time metrics is well-taken and directly addressed below.

read point-by-point responses

  1. Referee: [Abstract / real-world experiments] Abstract (hardware verification paragraph) and real-world experiments section: only traveled distance (up to 25% lower than FALCON) is reported for the Modal AI drone trials in two indoor environments; total exploration time, coverage completion time, or yaw-rotation overhead are omitted. This is load-bearing for the central claim because the abstract and simulation results explicitly note that yaw rotations increase total time relative to baselines, yet the practical net benefit (distance savings outweighing rotation duration) cannot be assessed without the corresponding time metrics in hardware.

    Authors: We agree that the absence of time metrics in the hardware section limits assessment of the net benefit, given the simulation results explicitly note increased total time from yaw rotations. The hardware trials prioritized verification of the distance savings (the paper's core path-efficiency claim) on the Modal AI platform, and quantitative timing data including rotation overhead was not recorded. We will revise the abstract's hardware paragraph and the real-world experiments section to explicitly state this limitation, clarify that distance is the primary reported metric, and note that time-distance tradeoffs were characterized via the tunable search radius in simulation. This revision will be made without altering the reported distance results. revision: yes

Circularity Check

0 steps flagged

No circularity; empirical validation against external baselines

full rationale

The paper presents OPAL as a new algorithmic framework for aerial exploration using deliberate yaw rotations at branch points, with variants using model-free, LLM, or VLM frontier selection. All performance claims (shorter travel distance, higher coverage AUC, 25% distance reduction vs FALCON) rest on direct simulation and hardware comparisons to independent external methods (EDEN, FALCON). No equations, fitted parameters, self-citations, or uniqueness theorems are invoked; the method is a design choice whose benefits are measured empirically rather than derived by construction from its own inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The method relies on standard domain assumptions about sensor visibility during yaw and environment structure; the only tunable element is the search radius. No new physical entities or complex fitted constants are introduced.

free parameters (1)
  • vicinity search radius
    Varied during experiments to control which frontiers are considered, creating an explicit tradeoff between travel distance and total time.
axioms (1)
  • domain assumption A 360-degree yaw rotation at branch points supplies sufficient visual data for improved frontier selection relative to global planning.
    This premise is invoked as the central mechanism distinguishing OPAL from baselines.

pith-pipeline@v0.9.1-grok · 5765 in / 1385 out tokens · 40130 ms · 2026-06-29T21:48:26.246449+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

48 extracted references · 3 canonical work pages · 1 internal anchor

  1. [1]

    A frontier-based approach for autonomous exploration,

    B. Yamauchi, “A frontier-based approach for autonomous exploration,” inProceedings of the 1997 IEEE International Symposium on Com- putational Intelligence in Robotics and Automation (CIRA), 1997, pp. 146–151

    1997

  2. [2]

    FALCON: Fast autonomous aerial exploration using coverage path guidance,

    Y . Zhang, X. Chen, C. Feng, B. Zhou, and S. Shen, “FALCON: Fast autonomous aerial exploration using coverage path guidance,”IEEE Transactions on Robotics, 2024

    2024

  3. [3]

    EDEN: Efficient dual-layer exploration planning for fast UA V autonomous ex- ploration in large 3-D environments,

    Q. Dong, X. Zhang, S. Zhang, Z. Wang, Z. Ma, and H. Xi, “EDEN: Efficient dual-layer exploration planning for fast UA V autonomous ex- ploration in large 3-D environments,”IEEE Transactions on Industrial Electronics, 2025

    2025

  4. [4]

    Lantern-explorer: A collision-avoidance autonomous exploration drone system based on laser slam with optimized hardware and software,

    L. Zhu, R. Zhong, D. Xie, and X. Yuan, “Lantern-explorer: A collision-avoidance autonomous exploration drone system based on laser slam with optimized hardware and software,”ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. X-G-2025, pp. 1077–1084, 2025

    2025

  5. [5]

    Autonomous frontier-based exploration with high-level vlm guidance,

    A. Aitha, “Autonomous frontier-based exploration with high-level vlm guidance,” Master’s thesis, University of California, Berkeley, August 2025, https://www-video.eecs.berkeley.edu/papers/aarushaitha/aarush thesis.pdf

    2025

  6. [6]

    Autonomous frontier-based exploration with high-level VLM guidance,

    A. Aitha and A. Zakhor, “Autonomous frontier-based exploration with high-level VLM guidance,” June 2026, accepted to the CVPR 2026 2nd Workshop on 3D-LLM/VLA: Bridging Language, Vision and Action in 3D Environments

    2026

  7. [7]

    Receding horizon “next-best-view

    A. Bircher, M. Kamel, K. Alexis, H. Oleynikova, and R. Siegwart, “Receding horizon “next-best-view” planner for 3D exploration,” in 2016 IEEE International Conference on Robotics and Automation (ICRA), 2016, pp. 1462–1468

    2016

  8. [8]

    Efficient autonomous exploration planning of large-scale 3-d environments,

    M. Selin, M. Tiger, D. Duberg, F. Heintz, and P. Jensfelt, “Efficient autonomous exploration planning of large-scale 3-d environments,” IEEE Robotics and Automation Letters, vol. 4, no. 2, pp. 1699–1706, 2019

    2019

  9. [9]

    Informed sampling exploration path planner for 3d reconstruction of large scenes,

    Y . Kompis, L. Bartolomei, R. Mascaro, L. Teixeira, and M. Chli, “Informed sampling exploration path planner for 3d reconstruction of large scenes,”IEEE Robotics and Automation Letters, vol. 6, no. 4, pp. 7894–7901, 2021

    2021

  10. [10]

    A shadowcasting-based next-best-view planner for autonomous 3d ex- ploration,

    A. Batinovic, A. Ivanovic, T. Petrovic, and S. Bogdan, “A shadowcasting-based next-best-view planner for autonomous 3d ex- ploration,”IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 2969–2976, 2022

    2022

  11. [11]

    Fuel: Fast uav exploration using incremental frontier structure and hierarchical planning,

    B. Zhou, Y . Zhang, X. Chen, and S. Shen, “Fuel: Fast uav exploration using incremental frontier structure and hierarchical planning,”IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 779–786, 2021

    2021

  12. [12]

    Fsmp: A frontier-sampling-mixed planner for fast autonomous exploration of complex and large 3-d environments,

    S. Zhang, X. Zhang, Q. Dong, Z. Wang, H. Xi, and J. Yuan, “Fsmp: A frontier-sampling-mixed planner for fast autonomous exploration of complex and large 3-d environments,”IEEE Transactions on Instrumentation and Measurement, vol. 74, pp. 1–14, 2025

    2025

  13. [13]

    Flare: Fast large-scale autonomous exploration guided by unknown regions,

    X. Liu, M. L. Lin, S. Li, G. Xu, Z. Wang, H. Wu, and Y . Liu, “Flare: Fast large-scale autonomous exploration guided by unknown regions,” IEEE Robotics and Automation Letters, vol. 10, no. 11, pp. 12 197– 12 204, 2025

    2025

  14. [14]

    Ptplanner: Efficient autonomous uav exploration via prior-enhanced and topology-aware hierarchical planning,

    C. Zhao, Z. Deng, Z. Zhang, and X. Guo, “Ptplanner: Efficient autonomous uav exploration via prior-enhanced and topology-aware hierarchical planning,”Drones, vol. 10, no. 3, p. 217, 2026

    2026

  15. [15]

    Receding horizon path planning for 3d exploration and surface inspection,

    A. Bircher, M. Kamel, K. Alexis, H. Oleynikova, and R. Siegwart, “Receding horizon path planning for 3d exploration and surface inspection,”Autonomous Robots, vol. 42, no. 2, pp. 291–306, 2018

    2018

  16. [16]

    Efficient autonomous exploration with incrementally built topological map in 3-d environments,

    C. Wang, H. Ma, W. Chen, L. Liu, and M. Q.-H. Meng, “Efficient autonomous exploration with incrementally built topological map in 3-d environments,”IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 12, pp. 9853–9865, 2020

    2020

  17. [17]

    Graph-based topological exploration planning in large-scale 3d environments,

    F. Yang, D.-H. Lee, J. Keller, and S. Scherer, “Graph-based topological exploration planning in large-scale 3d environments,” in2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, pp. 12 730–12 736

    2021

  18. [18]

    A multi-resolution frontier-based planner for autonomous 3d explo- ration,

    A. Batinovic, T. Petrovic, A. Ivanovic, F. Petric, and S. Bogdan, “A multi-resolution frontier-based planner for autonomous 3d explo- ration,”IEEE Robotics and Automation Letters, vol. 6, no. 3, pp. 4528– 4535, 2021

    2021

  19. [19]

    Tare: A hierarchical framework for efficiently exploring complex 3d environments,

    C. Cao, H. Zhu, H. Choset, and J. Zhang, “Tare: A hierarchical framework for efficiently exploring complex 3d environments,” in Proceedings of Robotics: Science and Systems (RSS), 2021

    2021

  20. [20]

    Autonomous UA V exploration of dynamic environments via incremental sampling and probabilistic roadmap,

    Z. Xu, D. Deng, and K. Shimada, “Autonomous UA V exploration of dynamic environments via incremental sampling and probabilistic roadmap,”IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1679–1686, 2021

    2021

  21. [21]

    Autonomous 3d exploration in large-scale environments with dynamic obstacles,

    E. Wiman, L. Widen, M. Tiger, and F. Heintz, “Autonomous 3d exploration in large-scale environments with dynamic obstacles,” in 2024 IEEE International Conference on Robotics and Automation (ICRA), 2024, pp. 2389–2395

    2024

  22. [22]

    Autonomous exploration method for fast unknown environment mapping by using uav equipped with limited fov sensor,

    Y . Zhao, L. Yan, H. Xie, J. Dai, and P. Wei, “Autonomous exploration method for fast unknown environment mapping by using uav equipped with limited fov sensor,”arXiv preprint arXiv:2302.02293, 2023

    work page arXiv 2023

  23. [23]

    Autonomous exploration of unknown 3d environments using a frontier-based collector strategy,

    I. D. Changoluisa Caiza, A. Milas, M. A. Montes Grova, F. J. Perez- Grau, and T. Petrovic, “Autonomous exploration of unknown 3d environments using a frontier-based collector strategy,” in2024 IEEE International Conference on Robotics and Automation (ICRA), 2024, pp. 13 566–13 572

    2024

  24. [24]

    Efficient uav exploration for large-scale 3d environments using low-memory map,

    J. Huang, Z. Fan, Z. Yan, P. Duan, R. Mei, and H. Cheng, “Efficient uav exploration for large-scale 3d environments using low-memory map,”Drones, vol. 8, no. 9, p. 443, 2024

    2024

  25. [25]

    Pipe planner: Pathwise information gain with map predictions for indoor robot exploration,

    S. Baek, B. Moon, S. Kim, M. Cao, C. Ho, S. Scherer, and J. H. Jeon, “Pipe planner: Pathwise information gain with map predictions for indoor robot exploration,” in2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Oct. 2025

    2025

  26. [26]

    Path planner for autonomous exploration of underground mines by aerial vehicles,

    C. Rubio-Sierra, D. Dominguez, J. Gonzalo, and A. Escapa, “Path planner for autonomous exploration of underground mines by aerial vehicles,”Sensors, vol. 20, no. 15, p. 4259, 2020

    2020

  27. [27]

    Optimal frontier-based au- tonomous exploration in unconstructed environment using RGB-D sensor,

    L. Lu, C. Redondo, and P. Campoy, “Optimal frontier-based au- tonomous exploration in unconstructed environment using RGB-D sensor,”Sensors, vol. 20, no. 22, p. 6507, 2020

    2020

  28. [28]

    Surface-driven next-best-view planning for exploration of large-scale 3d environments,

    G. Hardouin, F. Morbidi, J. Moras, J. Marzat, and E. M. Mouaddib, “Surface-driven next-best-view planning for exploration of large-scale 3d environments,” inIF AC-PapersOnLine, vol. 53, no. 2, 2020, pp. 15 501–15 507

    2020

  29. [29]

    Topological map-based autonomous exploration in large-scale scenes for unmanned vehicles,

    Z. Cao, Z. Du, and J. Yang, “Topological map-based autonomous exploration in large-scale scenes for unmanned vehicles,”Drones, vol. 8, no. 4, p. 124, 2024

    2024

  30. [30]

    Topological frontier- based exploration and map-building using semantic information,

    C. Gomez, A. C. Hernandez, and R. Barber, “Topological frontier- based exploration and map-building using semantic information,” Sensors, vol. 19, no. 20, p. 4595, 2019

    2019

  31. [31]

    Asep: An autonomous seman- tic exploration planner with object labeling,

    A. Milas, A. Ivanovic, and T. Petrovic, “Asep: An autonomous seman- tic exploration planner with object labeling,”IEEE Access, vol. 11, pp. 107 169–107 183, 2023

    2023

  32. [32]

    Do As I Can, Not As I Say: Grounding Language in Robotic Affordances

    M. Ahn, A. Brohan, N. Brown, Y . Chebotar, O. Cortes, B. David, C. Finn, C. Fu, K. Gopalakrishnan, K. Hausman, A. Herzog, D. Ho, J. Hsu, J. Ibarz, B. Ichter, A. Irpan, E. Jang, R. J. Ruano, K. Jeffrey, S. Jesmonth, N. J. Joshi, R. Julian, D. Kalashnikov, Y . Kuang, K. Lee, S. Levine, Y . Lu, L. Luu, C. Parada, and P. Pastor, “Do as I can, not as I say: Gr...

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  33. [33]

    PaLM- E: An embodied multimodal language model,

    D. Driess, F. Xia, M. S. M. Sajjadi, C. Lynch, A. Chowdhery, B. Ichter, A. Wahid, J. Tompson, Q. Vuong, T. Yu, W. Huang, Y . Chebotar, P. Sermanet, D. Duckworth, S. Levine, V . Vanhoucke, K. Hausman, M. Toussaint, K. Greff, A. Zeng, I. Mordatch, and P. Florence, “PaLM- E: An embodied multimodal language model,” inProceedings of the 40th International Conf...

    2023

  34. [34]

    Depth-first search and linear graph algorithms,

    R. Tarjan, “Depth-first search and linear graph algorithms,”SIAM Journal on Computing, vol. 1, no. 2, pp. 146–160, 1972

    1972

  35. [35]

    Sensor-based exploration for general robotic systems,

    L. Freda, G. Oriolo, and F. Vecchioli, “Sensor-based exploration for general robotic systems,” in2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, 2008, pp. 2157–2164

    2008

  36. [36]

    Exploring unstructured environment with frontier trees,

    R. Korb and A. Sch ¨ottl, “Exploring unstructured environment with frontier trees,”Journal of Intelligent & Robotic Systems, vol. 91, no. 2, pp. 617–628, 2018

    2018

  37. [37]

    Habitat-Matterport 3D dataset (HM3D): 1000 large-scale 3D environments for embodied AI,

    S. K. Ramakrishnan, A. Gokaslan, E. Wijmans, O. Maksymets, A. Clegg, J. M. Turner, E. Undersander, W. Galuba, A. Westbury, A. X. Chang, M. Savva, Y . Zhao, and D. Batra, “Habitat-Matterport 3D dataset (HM3D): 1000 large-scale 3D environments for embodied AI,” inAdvances in Neural Information Processing Systems Datasets and Benchmarks Track, 2021

    2021

  38. [38]

    Accessed May 19,

    Blender Foundation,Blender, Blender Foundation, Amsterdam, The Netherlands, 2026, open-source 3D creation suite. Accessed May 19,

    2026

  39. [39]

    Available: https://www.blender.org/

    [Online]. Available: https://www.blender.org/

  40. [40]

    Gleam: Learning generalizable exploration policy for active mapping in com- plex 3d indoor scenes,

    X. Chen, T. Wang, Q. Li, T. Huang, J. Pang, and T. Xue, “Gleam: Learning generalizable exploration policy for active mapping in com- plex 3d indoor scenes,”arXiv preprint arXiv:2505.20294, 2025

    work page arXiv 2025

  41. [41]

    Gennbv: Generalizable next-best-view policy for active 3d reconstruction,

    X. Chen, Q. Li, T. Wang, T. Xue, and J. Pang, “Gennbv: Generalizable next-best-view policy for active 3d reconstruction,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion (CVPR), June 2024, pp. 16 436–16 445

    2024

  42. [42]

    Hestia: V oxel-face-aware hierarchical next- best-view acquisition for efficient 3d reconstruction,

    C.-Y . Lu, Z. Zhuang, N. T. T. Le, D. Xiao, Y .-C. Chang, T. Do, S. Sridhar, and C.-T. Lin, “Hestia: V oxel-face-aware hierarchical next- best-view acquisition for efficient 3d reconstruction,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Mar. 2026, pp. 5302–5312

    2026

  43. [43]

    VOXL 2 - D0014 Architecture (Starling 2),

    ModalAI, “VOXL 2 - D0014 Architecture (Starling 2),” https://docs. modalai.com/voxl2-d0014/, 2026, accessed April 30, 2026

    2026

  44. [44]

    VOXL 2 Coax Camera Bundles,

    ——, “VOXL 2 Coax Camera Bundles,” https://docs.modalai.com/ voxl2-coax-camera-bundles/, 2026, accessed April 30, 2026

    2026

  45. [45]

    VOXL Open-VINS Server,

    ——, “VOXL Open-VINS Server,” https://docs.modalai.com/ voxl-open-vins-server/, 2026, accessed April 30, 2026

    2026

  46. [46]

    Modal Pipe Architecture,

    ——, “Modal Pipe Architecture,” https://docs.modalai.com/mpa/, 2026, accessed April 30, 2026

    2026

  47. [47]

    VOXL Vision Hub,

    ——, “VOXL Vision Hub,” https://docs.modalai.com/ voxl-vision-hub/, 2026, accessed April 30, 2026

    2026

  48. [48]

    VOXL PX4,

    ——, “VOXL PX4,” https://docs.modalai.com/voxl-px4/, 2026, ac- cessed April 30, 2026

    2026