DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells

H. Jin Kim; Jungwon Park; Yunwoo Lee

REVIEW 3 major objections 4 minor 1 cited by

DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that two time-varying cells—a Dynamic Buffered Voronoi Cell and a Dynamic Inter-Visibility Cell—make distributed multi-agent target tracking collision-free, occlusion-free, and computable in milliseconds.

desk verdict A genuinely new distributed-tracking construction with a real gap: the DIVC occlusion guarantee is asserted, not proved, and the paper needs a fix before its central claim is solid. read the letter →

arxiv 2411.18086 v2 pith:N7U7OXTC submitted 2024-11-27 cs.RO cs.SYeess.SY

Yunwoo Lee , Jungwon Park , H. Jin Kim This is my paper · ORCID

classification cs.ROcs.SYeess.SY

keywords distributedmulti-agentplanningtargettrackingocclusionavoidancecollisiondynamicVoronoicellsBernsteinpolynomialmotionprimitivesaerialroboticstrajectory

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 letter claims that two time-varying geometric cells can make distributed multi-agent target tracking both collision-free and occlusion-free in real time. The Dynamic Buffered Voronoi Cell (DBVC) keeps agents apart, while the Dynamic Inter-Visibility Cell (DIVC) keeps each agent's line of sight to the target clear of its teammates. Because the cells translate with the target's predicted motion, agents can plan independently in milliseconds, which the authors demonstrate in simulations and hardware flights with dozens of obstacles.

What carries the argument

The load-bearing objects are the Dynamic Buffered Voronoi Cell (DBVC), a time-shifting half-space that guarantees inter-agent separation, and the Dynamic Inter-Visibility Cell (DIVC), a pair of half-spaces that are supposed to keep each agent's view of the target unblocked by teammates. The DBVC generalizes the static Buffered Voronoi Cell by translating with the target's predicted displacement $x_q(t)-x_{q0}$, which prevents agents from getting stuck while following a moving target. The DIVC is built in two cases depending on whether the angle between the two agents' lines of sight is obtuse or acute; in each case the half-spaces are defined so that, for a parameter $\alpha$ within the allowed range, any point in their intersection satisfies the line-of-sight clearance inequality (11). Bernstein polynomial motion primitives carry the trajectories, and their convex hull property lets the planner verify inclusion in these affine cells by checking control points, while their integral property speeds up the selection of the best primitive.

What would settle it

Numerically test inequality (11) for a dense grid of agent and target positions near the boundary of the allowed $\alpha$ ranges (10) and (14); if any point in the DIVC intersection lies within $r_c$ of the opposite line-of-sight segment, the core occlusion guarantee is false. A simpler experiment is to run the planner with two agents in an empty environment and measure the minimum distance between each line of sight and the other agent over time; a value below $r_c$ falsifies the claim.

Watch

Extended reading notes

Core claim

The central claim is that the DBVC and the DIVC are time-variant spaces that prevent mutual collisions and occlusions among agents while enabling them to maintain suitable distances from the moving target. The DBVC is a half-space built from the agents' current positions and the target's predicted trajectory, and Lemma 1 shows that if each of two agents lies in its own DBVC half-space, their distance is at least $2r_c$. The DIVC is the intersection of two half-spaces per agent pair, constructed separately for obtuse and acute viewing angles, and the paper asserts that any trajectory inside them keeps the segment from the agent to the target at distance greater than $r_c$ from every other agent. These cells are combined with Bernstein polynomial motion primitives and a sample-check-select strategy; the convex hull and integral properties of Bernstein polynomials make the feasibility checks fast, and a refined, less conservative check than the authors' previous work increases the success rate. The algorithm computes each agent's trajectory within several milliseconds on an Intel i7 desktop.

Load-bearing premise

The paper's occlusion-avoidance guarantee depends on an unproved geometric assertion: that for any $\alpha$ in the stated range, every point in the DIVC intersection of two half-spaces lies at distance greater than $r_c$ from the line segment joining the other agent to the target; if that assertion fails for some configuration, the method can produce trajectories that block a teammate's view even though the planner ran exactly as specified.

Editorial extensions

If this is right

Each agent can replan independently from shared positions and the predicted target trajectory, without priority ordering or consensus among teammates.
The time-varying cells avoid the deadlock that static buffered Voronoi cells cause when the target moves, keeping agents within the commanded tracking distance.
The relaxed feasibility check expands the set of admissible primitives, which the benchmark shows as higher success rates in dynamic obstacle fields.
Reported per-agent computation under 10 ms at 1000 sampled primitives supports real-time replanning on small onboard computers.
In the comparison, the DBVC and DIVC raise success rates dramatically as the number of trackers grows, especially at short tracking distances.

Reading between the lines

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

A proof or counterexample for inequality (11) would settle whether the occlusion-free guarantee is absolute; the simulations alone cannot distinguish a valid cell construction from a plannable one that occasionally fails.
The cell construction assumes synchronized replanning and shared positions; asynchronous communication or latency would break the geometric assumptions, so the method's claims are strongest for tightly synchronized teams.
The same translated-cell idea could be extended to shape-changing cells (the authors list this as future work) or to heterogeneous agent radii, which would let the planner handle mixed fleets.
A direct extension of the DIVC to 3D would require the obtuse/acute construction to be replaced by a spherical or conic condition, so the 3D demonstrations shown here are empirical rather than guaranteed by the lemmas.

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.

Desk Editor's Note

read the letter

Short version: useful paper, real new idea, but the load-bearing occlusion-avoidance claim is not actually proven. The DBVC and DIVC are time-varying cells that move with the target, a genuine step beyond the static BVC and the authors' single-agent BPMP tracker. Lemma 1 for collision avoidance is fine. The less conservative Bernstein-based feasibility check is a real improvement, and the validation is extensive: simulations with dozens of obstacles, Crazyflie hardware, and success-rate comparisons.

The soft spot is Section IV-B. Inequality (11) is the entire bridge from the DIVC half-space constraints to 'no agent occludes another agent's line of sight.' The paper states it as 'the following inequalities are satisfied' and moves on. Lemma 2 only shows the cells are nonempty by translating with the target; it does not establish (11). The alpha bounds in (10)/(14) are justified by geometric claims, but no derivation is given. That matters: the strict form fails at equality in at least one concrete configuration (the stress-test example), and the paper never tells you how alpha is chosen inside the feasible range, so equality cases are not ruled out. This gap is not fatal if patched; the rest of the paper is assembled carefully, and I would expect a proof to fit in an appendix. Right now, the main guarantee rests on an unproved lemma.

Smaller issues: the 3D extension is only examples, no mathematical generalization—fine for a letter as long as it is described as heuristic. Table II has at least one row that looks internally inconsistent (Sc1 proposed row gives agent-to-agent distance min/mean 0.561/0.727 and line-of-sight-to-agent distance min/mean 0.290/0.727; the identical means with different minima are at least suspicious). No code or parameters are released, which makes the success-rate tables hard to audit.

For whom: anyone in multi-drone tracking or distributed trajectory generation should read this; it is within-subfield significant. It deserves a serious referee. I would send it to review with the expectation that the DIVC proof appears or the claim is softened. The central idea is plausible and the flaw is addressable.

Referee Report

3 major / 4 minor

Summary. The paper proposes a distributed trajectory planner for multi-agent aerial target tracking. The key components are the Dynamic Buffered Voronoi Cell (DBVC) for inter-agent collision avoidance and the Dynamic Inter-Visibility Cell (DIVC) for inter-agent occlusion avoidance. These time-varying cells are combined with Bernstein-polynomial motion primitives and a sample-check-select strategy. The authors claim collision- and occlusion-free tracking with millisecond computation times, validated in simulations, hardware experiments, and two 3D scenarios.

Significance. If the central geometric claims are correct, the work has clear practical value: it offers a fully distributed, non-priority-based method for multi-agent tracking that simultaneously addresses collision, occlusion, and target-distance constraints, with a less conservative feasibility check than the authors' prior work. The DBVC Lemma 1 is proved, the use of Bernstein convex-hull properties is sound, and the empirical validation is extensive, including hardware demonstrations and comparisons against noncooperative and conservative baselines. The main unresolved issue is that the DIVC occlusion-avoidance property, which is the load-bearing theoretical bridge between the half-space constraints and the no-occlusion guarantee, is asserted without proof.

major comments (3)

[Section IV-B, Eqs. (10)-(16) and Eq. (20c)] If the authors cannot provide such a proof, the claim that DIVC prevents inter-agent occlusion is unsupported, and the validation results would not justify the proposed method's main contribution.
[Section VI-D, 3D Extensions] As written, the 3D experimental results cannot be interpreted as evidence for the same collision- and occlusion-free guarantee that is claimed for the 2D case.
[Table II, Scenario 1 row] The authors should also verify that the same ordering convention is used across all rows and columns of the table.

minor comments (4)

[Section IV-B, acute case] This notational inconsistency makes the construction harder to follow.
[Section V-B-2, Eq. (21)] Please correct the subscript for consistency with the other sigma terms.
[Section IV-A, Lemma 2] Clarify that both agents translate with the target (or otherwise specify what is assumed about the other agent's motion) so that the non-emptiness argument is complete.
[Section VI-A, Scenario 1] This is a minor language issue that does not affect the technical content.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the central derivation is self-contained, though the DIVC visibility lemma (11) is asserted without proof and should be treated as a correctness gap.

full rationale

The paper's derivation chain is not circular. The DBVC in (7) is a time-translated buffered Voronoi cell, and Lemma 1 derives the collision-free separation from the two half-space inequalities via Cauchy-Schwarz, which is an independent geometric argument. The DIVC in (12) and (15) defines half-spaces from the target-agent geometry and then asserts the visibility separation (11); even though (11) is load-bearing for the central occlusion-avoidance claim and is not proved, it is not equivalent to the cell definitions by construction, so it is a missing lemma rather than a circular reduction. The planner's feasibility checks (19)-(23) are sufficient conditions stated directly in the paper, with the Bernstein convex-hull arguments used to verify them, and the claimed avoidance properties are not fitted parameters renamed as outputs. Self-citations to [8] supply target prediction, safe corridors, and a conservative baseline, but those components are external to the new DBVC/DIVC claim and are used as tools or comparison points, not as the proof of the central result. The claimed improvement over [8] is substantiated by the displayed sigma-prime inequalities (21) and by comparative success-rate tables. Therefore no fitted input is called a prediction, no uniqueness result is imported from the authors, and no known result is merely renamed. The omitted derivation of inequality (11) should be weighed as a correctness risk in review rather than as circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 2 invented entities

The central claim rests on the two new cell types: the DBVC has a proof (Lemma 1) while the DIVC's key property is unproved. Several tuning parameters (alpha, dmax, wj) are unreported. The method assumes synchronized, shared-state execution and accurate target prediction.

free parameters (4)

alpha_o and alpha_a
Per-pair scaling factors in the DIVC construction (Eqs. 10 and 14). The paper requires a range but does not report the concrete values used in experiments; the sizes of the visibility cells depend on this choice.
d_des desired distance = 0.45 m (midpoint of sampling range (0.3, 0.6) m)
Desired tracking distance in the cost (24c) is set to half of the sum of the sampling range bounds; a tuning choice.
wj jerk weight
Weight in the cost (24a); no value is reported.
dmin and dmax distance bounds = dmin set to rq + rc; dmax unreported
Distance constraints (5) require both bounds; only dmin is explicitly set in validation.

assumptions (5)

ad hoc to paper The DIVC half-spaces, for any alpha in the stated range, guarantee the line-of-sight distance inequalities (11) for all points in the cells.
This is the core geometric claim of Section IV-B, stated without a complete proof. It is specific to this paper's construction.
domain assumption All agents share current positions and start trajectory planning at the same time.
Assumption in Section III-C; the distributed guarantees rest on synchronized execution and shared state.
domain assumption Agent trajectories are represented as Bernstein polynomials and the quadrotor dynamics are differentially flat.
Standard in the field; used throughout Sections IV and V.
domain assumption Target and obstacle future trajectories are predicted by the method in [8] and a constant velocity model, and the predictions are accurate enough for the time-varying cells to remain valid.
The cells translate with the predicted target trajectory; poor prediction can break the guarantees. Stated in Section III-C.
standard math Convex hull and integral properties of Bernstein polynomials hold for the coefficient checks.
Used in feasibility checks (19), (20), and cost selection (24).

invented entities (2)

Dynamic Buffered Voronoi Cell (DBVC)
purpose: Time-varying half-space that prevents inter-agent collisions while tracking the target.
A new geometric construct built from BVC [7] plus target displacement; its validity rests on Lemma 1 and the paper's assumptions, not on external evidence.
Dynamic Inter-Visibility Cell (DIVC)
purpose: Time-varying half-space set intended to prevent inter-agent occlusion of the target.
A new construct; the key occlusion-avoidance property (11) is asserted without a full proof, so it is not independently established.

how reviews work

0 comments

Cite this review

Pith. "Pith review of DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells." pith.science (2026). https://pith.science/paper/N7U7OXTC

@misc{pith2026241118086,
  author       = {Pith},
  title        = {Pith review of: DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N7U7OXTC}},
  note         = {Machine review of arXiv:2411.18086}
}

read the original abstract

This letter presents a distributed trajectory planning method for multi-agent aerial tracking. The proposed method uses a Dynamic Buffered Voronoi Cell (DBVC) and a Dynamic Inter-Visibility Cell (DIVC) to formulate the distributed trajectory generation. Specifically, the DBVC and the DIVC are time-variant spaces that prevent mutual collisions and occlusions among agents, while enabling them to maintain suitable distances from the moving target. We combine the DBVC and the DIVC with an efficient Bernstein polynomial motion primitive-based tracking generation method, which has been refined into a less conservative approach than in our previous work. The proposed algorithm can compute each agent's trajectory within several milliseconds on an Intel i7 desktop. We validate the tracking performance in challenging scenarios, including environments with dozens of obstacles.

Figures

Figures reproduced from arXiv: 2411.18086 by the authors.

**Figure 2.** DIVC fomulation the intersections between the segments and lines are denoted as Vaj and Vai, respectively. Then, we draw the rays (blue in Fig. 2b) starting from Vai and Vaj , which are parallel to the lines that pass through VQ and are tangential to the balls B(Vaj , rc) and B(Vai, rc). Half-spaces divided by the rays are represented as H ij o2 and H ji o2 . To satisfy conditions that Vai and Vaj are on ViVQ and Vj… view at source ↗

**Figure 3.** Comparison of the feasibility checks. Yellow: an area where [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 5.** Figure 5: Target tracking experiments. The total flight paths of [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Tracking in 3D spaces. Reported paths of trackers (blue), target [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

Visibility-Aware Cooperative Tracking with Decentralized LiDAR-Based Aerial Swarms
cs.RO 2025-12 conditional novelty 6.0 of 10

A decentralized LiDAR-swarm planner that maintains near-perfect target visibility by encoding 3-D occlusion as a Spherical Signed Distance Field and coordinating 3-D encirclement via an electrostatic-potential-inspired cost.

Reference graph

Works this paper leans on

21 extracted references · 19 canonical work pages · cited by 1 Pith paper

[1]

Do you see what i see? coordinating multiple aerial cameras for robot cinematography,

A. Bucker, R. Bonatti, and S. Scherer, “Do you see what i see? coordinating multiple aerial cameras for robot cinematography,” in 2021 IEEE International Conference on Robotics and Automation (ICRA) , 2021, pp. 7972–7979

work page 2021
[2]

Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,

A. Alc ´antara, J. Capit ´an, R. Cunha, and A. Ollero, “Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,” Robotics and Autonomous Systems , vol. 140, p. 103778, 2021

work page 2021
[3]

Autonomous aerial filming with distributed lighting by a team of unmanned aerial vehicles,

V . Kr´atk´y, A. Alc ´antara, J. Capit ´an, P. ˇStˇep´an, M. Saska, and A. Ollero, “Autonomous aerial filming with distributed lighting by a team of unmanned aerial vehicles,” IEEE Robotics and Automation Letters , vol. 6, no. 4, pp. 7580–7587, 2021

work page 2021
[4]

Active perception based formation control for multiple aerial vehicles,

R. Tallamraju, E. Price, R. Ludwig, K. Karlapalem, H. H. B ¨ulthoff, M. J. Black, and A. Ahmad, “Active perception based formation control for multiple aerial vehicles,” IEEE Robotics and Automation Letters , vol. 4, no. 4, pp. 4491–4498, 2019

work page 2019
[5]

3d human reconstruction in the wild with collaborative aerial cameras,

C. Ho, A. Jong, H. Freeman, R. Rao, R. Bonatti, and S. Scherer, “3d human reconstruction in the wild with collaborative aerial cameras,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 5263–5269

work page 2021
[6]

Markerless motion capture for humans through a multi-uav system,

S. Zhuge, Y . He, X. Xu, S. Gan, C. Li, B. Lin, X. Yang, and X. Zhang, “Markerless motion capture for humans through a multi-uav system,” IEEE Transactions on Instrumentation and Measurement , 2023

work page 2023
[7]

Fast, on-line collision avoidance for dynamic vehicles using buffered voronoi cells,

D. Zhou, Z. Wang, S. Bandyopadhyay, and M. Schwager, “Fast, on-line collision avoidance for dynamic vehicles using buffered voronoi cells,” IEEE Robotics and Automation Letters , vol. 2, no. 2, pp. 1047–1054, 2017

work page 2017
[8]

Bpmp-tracker: A versatile aerial target tracker using bernstein polynomial motion primitives,

Y . Lee, J. Park, B. Jeon, S. Jung, and H. J. Kim, “Bpmp-tracker: A versatile aerial target tracker using bernstein polynomial motion primitives,” IEEE Robotics and Automation Letters , pp. 1–8, 2024

work page 2024

Show all 21 references

[9]

Elastic tracker: A spatio-temporal trajectory planner for flexible aerial tracking,

J. Ji, N. Pan, C. Xu, and F. Gao, “Elastic tracker: A spatio-temporal trajectory planner for flexible aerial tracking,” in 2022 International Conference on Robotics and Automation (ICRA) , 2022, pp. 47–53

2022
[10]

Auto filmer: Autonomous aerial videography under human interaction,

Z. Zhang, Y . Zhong, J. Guo, Q. Wang, C. Xu, and F. Gao, “Auto filmer: Autonomous aerial videography under human interaction,” IEEE Robotics and Automation Letters , vol. 8, no. 2, pp. 784–791, 2023

2023
[11]

Integrated motion planner for real-time aerial videography with a drone in a dense environment,

B. Jeon, Y . Lee, and H. J. Kim, “Integrated motion planner for real-time aerial videography with a drone in a dense environment,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) , 2020, pp. 1243–1249

2020
[12]

Real- time motion planning for aerial videography with dynamic obstacle avoidance and viewpoint optimization,

T. N ¨ageli, J. Alonso-Mora, A. Domahidi, D. Rus, and O. Hilliges, “Real- time motion planning for aerial videography with dynamic obstacle avoidance and viewpoint optimization,” IEEE Robotics and Automation Letters, vol. 2, no. 3, pp. 1696–1703, 2017

2017
[13]

Real-time multi-convex model predictive control for occlusion-free target tracking with quadrotors,

H. Masnavi, V . K. Adajania, K. Kruusam¨ae, and A. K. Singh, “Real-time multi-convex model predictive control for occlusion-free target tracking with quadrotors,” IEEE Access , vol. 10, pp. 29 009–29 031, 2022

2022
[14]

Qp chaser: Polynomial trajectory generation for autonomous aerial tracking,

Y . Lee, J. Park, S. Jung, B. F. Jeon, D. Oh, and H. J. Kim, “Qp chaser: Polynomial trajectory generation for autonomous aerial tracking,” ArXiv, vol. abs/2302.14273, 2023

2023
[15]

Real-time planning for automated multi-view drone cinematography,

T. N ¨ageli, L. Meier, A. Domahidi, J. Alonso-Mora, and O. Hilliges, “Real-time planning for automated multi-view drone cinematography,” ACM Transactions on Graphics (TOG) , vol. 36, no. 4, pp. 1–10, 2017

2017
[16]

Autonomous drone cinematographer: Using artistic principles to create smooth, safe, occlusion-free trajectories for aerial filming,

R. Bonatti, Y . Zhang, S. Choudhury, W. Wang, and S. A. Scherer, “Autonomous drone cinematographer: Using artistic principles to create smooth, safe, occlusion-free trajectories for aerial filming,” in International Symposium on Experimental Robotics , 2018. [Online]. Availabl...

2018
[17]

Vision-based reactive planning for aggressive target tracking while avoiding collisions and occlusions,

B. Penin, P. R. Giordano, and F. Chaumette, “Vision-based reactive planning for aggressive target tracking while avoiding collisions and occlusions,” IEEE Robotics and Automation Letters , vol. 3, no. 4, pp. 3725–3732, 2018

2018
[18]

Robust multi-robot active target tracking against sensing and communication attacks,

L. Zhou and V . Kumar, “Robust multi-robot active target tracking against sensing and communication attacks,” IEEE Transactions on Robotics , vol. 39, no. 3, pp. 1768–1780, 2023

2023
[19]

Resilient active target tracking with multiple robots,

L. Zhou, V . Tzoumas, G. J. Pappas, and P. Tokekar, “Resilient active target tracking with multiple robots,” IEEE Robotics and Automation Letters, vol. 4, no. 1, pp. 129–136, 2019

2019
[20]

Swarm of micro flying robots in the wild,

X. Zhou, X. Wen, Z. Wang, Y . Gao, H. Li, Q. Wang, T. Yang, H. Lu, Y . Cao, C. Xu, and F. Gao, “Swarm of micro flying robots in the wild,” Science Robotics , vol. 7, 05 2022

2022
[21]

Dlsc: Distributed multi-agent trajectory planning in maze-like dynamic environments using linear safe corridor,

J. Park, Y . Lee, I. Jang, and H. J. Kim, “Dlsc: Distributed multi-agent trajectory planning in maze-like dynamic environments using linear safe corridor,” IEEE Transactions on Robotics, vol. 39, no. 5, pp. 3739–3758, 2023

2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Do you see what i see? coordinating multiple aerial cameras for robot cinematography,

A. Bucker, R. Bonatti, and S. Scherer, “Do you see what i see? coordinating multiple aerial cameras for robot cinematography,” in 2021 IEEE International Conference on Robotics and Automation (ICRA) , 2021, pp. 7972–7979

work page 2021

[2] [2]

Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,

A. Alc ´antara, J. Capit ´an, R. Cunha, and A. Ollero, “Optimal trajectory planning for cinematography with multiple unmanned aerial vehicles,” Robotics and Autonomous Systems , vol. 140, p. 103778, 2021

work page 2021

[3] [3]

Autonomous aerial filming with distributed lighting by a team of unmanned aerial vehicles,

V . Kr´atk´y, A. Alc ´antara, J. Capit ´an, P. ˇStˇep´an, M. Saska, and A. Ollero, “Autonomous aerial filming with distributed lighting by a team of unmanned aerial vehicles,” IEEE Robotics and Automation Letters , vol. 6, no. 4, pp. 7580–7587, 2021

work page 2021

[4] [4]

Active perception based formation control for multiple aerial vehicles,

R. Tallamraju, E. Price, R. Ludwig, K. Karlapalem, H. H. B ¨ulthoff, M. J. Black, and A. Ahmad, “Active perception based formation control for multiple aerial vehicles,” IEEE Robotics and Automation Letters , vol. 4, no. 4, pp. 4491–4498, 2019

work page 2019

[5] [5]

3d human reconstruction in the wild with collaborative aerial cameras,

C. Ho, A. Jong, H. Freeman, R. Rao, R. Bonatti, and S. Scherer, “3d human reconstruction in the wild with collaborative aerial cameras,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 5263–5269

work page 2021

[6] [6]

Markerless motion capture for humans through a multi-uav system,

S. Zhuge, Y . He, X. Xu, S. Gan, C. Li, B. Lin, X. Yang, and X. Zhang, “Markerless motion capture for humans through a multi-uav system,” IEEE Transactions on Instrumentation and Measurement , 2023

work page 2023

[7] [7]

Fast, on-line collision avoidance for dynamic vehicles using buffered voronoi cells,

D. Zhou, Z. Wang, S. Bandyopadhyay, and M. Schwager, “Fast, on-line collision avoidance for dynamic vehicles using buffered voronoi cells,” IEEE Robotics and Automation Letters , vol. 2, no. 2, pp. 1047–1054, 2017

work page 2017

[8] [8]

Bpmp-tracker: A versatile aerial target tracker using bernstein polynomial motion primitives,

Y . Lee, J. Park, B. Jeon, S. Jung, and H. J. Kim, “Bpmp-tracker: A versatile aerial target tracker using bernstein polynomial motion primitives,” IEEE Robotics and Automation Letters , pp. 1–8, 2024

work page 2024