Pith. sign in

REVIEW 6 major objections 8 minor 40 references

Attention-based Learning for 3D Informative Path Planning

T0 review · 6 major / 8 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper argues that an attention-based policy trained with PPO can make an aerial robot adaptively choose 3D measurement locations that reduce Gaussian-process uncertainty in regions of interest faster than MCTS, CMA-ES, random…

desk verdict A solid incremental extension of CAtNIPP to 3D altitude-dependent sensing; the results are plausible but the evaluation is too thin to support the 'significant' claim as stated. read the letter →

arxiv 2506.08434 v1 pith:ZBWAPWGA submitted 2025-06-10 cs.RO

Rui Zhao , Xingjian Zhang , Yuhong Cao , Yizhuo Wang , Guillaume Sartoretti This is my paper
classification cs.RO
keywords informativepathplanningdeepreinforcementlearningattentionmechanismGaussianprocess3DUAVexplorationadaptivesensingaltitude-awareregionsofinterest
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

Adaptive informative path planning asks an aerial robot to choose where to measure next so that, within a fixed travel budget, its belief about an underlying field—such as weed patches or gas plumes—becomes as accurate as possible. This paper argues that a policy network built on attention over a graph of candidate 3D measurement positions can learn that sequential choice directly, including the altitude-dependent trade-off between a wider sensing footprint and noisier readings. The authors train the policy with PPO, rewarding reductions in posterior uncertainty inside regions of interest, and report that it reduces both map uncertainty and ground-truth reconstruction error faster than Monte Carlo tree search, CMA-ES, random sampling, and coverage paths, while replanning in milliseconds. They also show the trained policy transfers from smaller to larger maps without retraining. If the claim holds, attention-based RL offers a practical, real-time way to run adaptive 3D surveys with a single downward-facing camera.

What carries the argument

The load-bearing object is the attention-based encoder-decoder over the augmented graph G'. The encoder embeds each node with its coordinates and Laplacian eigenvector positional encoding, then applies a Transformer self-attention layer so that every node's Gaussian-process estimate (mean and variance within the sensor's field of view) can inform every other node's representation; the decoder uses a pointer-network-style cross-attention to produce a distribution over the current node's neighbors, which lets the policy handle a varying number of neighbors. The reward that drives the whole system is the normalized trace reduction of the posterior covariance inside the region of interest, and the policy is optimized with PPO. This combination is what lets the agent implicitly estimate the value of future observations and balance altitude, sensing footprint, and measurement accuracy.

What would settle it

Run the same trained policy and the MCTS and CMA-ES baselines against ground-truth fields that are not samples from the fitted Gaussian process (for example, fields with sharp discontinuities or heavy-tailed anomalies), and compare RMSE inside the region of interest; if the attention policy no longer achieves the fastest uncertainty and RMSE reduction at matched budgets, the central claim is refuted. Alternatively, perturb the GP hyperparameters used by the belief updater and check whether the learned policy's advantage over baselines collapses.

Watch

Extended reading notes

Core claim

The central claim is that treating the 3D informative path planning problem as a sequential decision problem over a probabilistic roadmap, and solving it with an attention-based policy, produces an adaptive planner that outperforms existing planners at reconstructing regions of interest under time constraints. The agent's observation is the current graph of candidate nodes enriched with the Gaussian-process mean and variance associated with each node; the policy selects the next neighbor to visit; and the reward is the normalized reduction of the trace of the posterior covariance within the region of interest. The paper's experiments show the policy drops uncertainty below 50 within about 100 seconds in a 15 by 15 field, reducing uncertainty by 86.43% and RMSE by 61.52% at a 200-second budget, against 72.67% and 58.53% for MCTS, with a planning runtime of about 5 milliseconds versus nearly 4 seconds. The same policy, trained only on 15 by 15 maps, is reported to remain the best performer on 20 by 20 maps and to beat CMA-ES in a Gazebo weed-detection simulation.

Load-bearing premise

The planner's advantage rests on the Gaussian-process belief being a faithful model of the true field, with kernel hyperparameters and the region-of-interest threshold fixed as fitted; if a real environment's spatial structure departs from that GP, the learned policy's superiority may not transfer.

Editorial extensions

If this is right

  • A single trained policy can replace expensive online search: replanning takes about 5 milliseconds versus about 4 seconds for MCTS at comparable or better uncertainty reduction.
  • Altitude becomes an explicit decision variable: the policy first surveys at 14 meters to lower global uncertainty, then descends to 8 meters to refine high-value areas.
  • The policy transfers across map sizes without retraining, at least from 15 by 15 to 20 by 20 grids.
  • The approach works with downward-facing image-like sensors rather than point probes, so it applies directly to camera-equipped UAVs.
  • In simulated deployment, the uncertainty trace drops faster than CMA-ES under a 200-second budget.

Reading between the lines

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

  • A natural extension is testing on nonstationary or non-Gaussian fields; because both the reward and the evaluation metrics are defined through the Gaussian-process belief, the planner's margin over baselines may shrink when the true field is poorly described by the assumed kernel.
  • The graph representation could support multi-robot settings by treating teammates as additional nodes or by sharing observations across agents, although the paper only lists multi-agent coordination as future work.
  • The generalization claim was tested only from 15 by 15 to 20 by 20 grids, so whether the policy extends to much larger or differently shaped domains remains an open question.
  • A useful diagnostic would be to ablate the Transformer attention layer, replacing it with a simpler graph convolutional encoder, to measure how much of the advantage comes from global attention versus the graph encoding and reward shaping.
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. This paper proposes an attention-based deep reinforcement learning (PPO) method for adaptive three-dimensional informative path planning. The environment is discretized as a probabilistic roadmap graph with nodes at two altitude levels; each node stores Gaussian-process belief statistics. A Transformer encoder with Laplacian positional encoding and a pointer-network decoder outputs a distribution over neighboring nodes. The reward is the normalized reduction of the GP posterior trace over a region of interest (ROI), scaled by a constant factor. The authors compare against MCTS, random, coverage, and CMA-ES baselines in a simulated 15x15 grid and a Gazebo weed-detection scenario, reporting uncertainty and RMSE reductions as well as runtime, and claim training-free generalization to larger maps.

Significance. If supported, the contribution is relevant: most learned IPP policies are limited to 2D or point-sensing, whereas altitude-aware image-like sensing requires explicitly trading footprint against measurement accuracy. The architecture is a reasonable adaptation of attention and pointer-network ideas to graph-based 3D IPP, and the reported runtime advantage over MCTS is practically attractive. The manuscript also includes both offline and Gazebo evaluations. However, the evidence as presented is not yet sufficient to establish the headline claims: the primary uncertainty metric is the same quantity used as the training reward, the RMSE margins over strong baselines are small, the number of trials is very low, no significance tests are reported, and the closest 3D DRL baseline is not compared.

major comments (6)
  1. [Eq. (5), Table I, Fig. 4] The uncertainty-reduction metric in Table I and Fig. 4 is the same quantity as the reward in Eq. (5): both are the normalized reduction of the GP posterior trace over the ROI. Consequently, the uncertainty columns mainly show that the policy learned to perform the GP/Kalman update, and they do not independently validate the belief quality. The RMSE columns are the only ground-truth check, and there the margins are modest (e.g., 61.52% vs 58.53% at 200 s offline; 67.29% vs 64.83% in simulated deployment) and are based on only four trials with no significance tests. Please report RMSE as the primary independent evidence, add confidence intervals or significance tests, and temper the wording 'significantly reduces uncertainty' accordingly.
  2. [V-A] The ground-truth maps are generated as independent Bernoulli random cells, while the belief model assumes a GP with kernel hyperparameters fitted on 'independent maps' (length scale 3.67, signal variance 1.82, noise variance 1.42). The evaluation therefore does not test the assumed generative model, and it remains unclear whether the learned policy's uncertainty reduction transfers to fields that actually follow the GP or to real fields. Please evaluate on GP draws using the fitted kernel, and on real or more realistic field data if possible, and report sensitivity to hyperparameter misspecification and to the sensor noise parameters a and b.
  3. [II, V-B] The closest prior 3D DRL baseline, Vashisth et al. (Ref. [27]), is discussed in Related Work but never compared experimentally. The exclusion of Ref. [9] is also questionable, since that reference is itself a DRL-based IPP method, not an MCTS-in-2D method as stated in Section V-B. Add a direct comparison to at least Ref. [27] (and ideally Ref. [9]) under the same evaluation protocol; otherwise the claim of outperforming state-of-the-art planners is not established. Note also that the offline and simulated settings use different baseline sets (MCTS/random/coverage versus CMA-ES), which complicates interpretation.
  4. [V-B, Table I] Only four trials are used for each condition, no significance tests or confidence intervals are reported, Table I contains no error bars, and there is no evidence that training was repeated across different seeds. Given the small RMSE differences, the statement that the approach 'significantly reduces' uncertainty and RMSE is not statistically supported. Please provide more trials, significance testing, and multi-seed training results, and show error bars or confidence intervals in all quantitative figures and tables.
  5. [III-A, V-A, VI-D] The ROI definition is inconsistent and underspecified. Eq. (XI) defines ROIs from the belief via μ_i + βσ_i ≥ μth, but Section V-A says the ground-truth field maps are 'randomly split into high- and low-value regions to define regions of interest,' and the Gazebo experiment uses μth=0.05 while the main evaluation uses μth=0.4. The value of β is never given, and the sensor model constants a, b and the FoV scaling rule are not specified. Because both the reward and the evaluation metrics depend on the ROI set, the experiments are not reproducible without clarifying which ROI definition is used for reward versus evaluation and for each experiment.
  6. [V-C] The generalization claim rests on a single larger map size (20×20) and is made without any quantitative results in the text or table. Please provide quantitative comparisons, including baseline comparisons, at multiple map sizes and across other distribution shifts such as budgets, altitude levels, start positions, and sensor parameters, to support the claim of training-free generalization.
minor comments (8)
  1. [III-A] Equation (1) presents a static optimization objective, but the paper's contribution is adaptive planning; please clarify how adaptivity is incorporated into the formal problem statement.
  2. [III-B, III-C] The symbol K is used for both the covariance kernel in Eq. (2) and the Kalman gain in Eq. (3); please use distinct notation to avoid confusion.
  3. [V-B] The text states 'A has 450 actions' but A is not defined, and the policy actually chooses among k=20 neighboring nodes at each step; please clarify what A denotes and how it relates to the action space.
  4. [III-C, V-A] The description of how each camera image is converted into per-cell GP measurements, and how the measurement noise matrix R in Eq. (3) is constructed, is missing; please specify or cite the exact measurement model.
  5. [V-B] The runtime advantage is described as a 'tenfold speedup' relative to MCTS, but the random and coverage baselines run in 0.0001 s; please clarify the hardware used, whether the reported runtime is per decision step or per full mission, and compare inference time on the same platform.
  6. [V-B] The coverage baseline is restricted to a fixed 8 m altitude, which may handicap it in a 3D IPP comparison; if a coverage baseline is included, it should be allowed to select altitudes as well.
  7. [V-A] The inverse sensor model is attributed to 'Popovic et al. [9]', but Ref. [9] is Ruckin et al.; please correct or verify the citation.
  8. [IV-B, IV-C] No ablation study is reported for the attention encoder, Laplacian positional encoding, pointer decoder, or reward scaling; at least one ablation would help identify which component drives the reported performance.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the reward/evaluation alignment is objective alignment rather than a hidden fit, and the self-citations are not load-bearing.

full rationale

The paper's central comparative claims are tested on new random maps against external baselines (MCTS, CMA-ES, random, coverage), so the uncertainty-reduction result is an empirical optimization outcome rather than a value forced by construction. The reward in Eq. (5) and the reported 'uncertainty reduction' in Table I and Fig. 4 are both trace reductions of the GP posterior over the ROI, but this is deliberate objective-evaluation alignment standard in IPP, not a circular step: all baselines are scored on the same metric, and the learned policy could have failed to optimize it. The RMSE metric provides an independent ground-truth check, and the GP hyperparameters (length scale 3.67, signal variance 1.82, noise variance 1.42) are fitted once on independent maps, not to the test outputs. The only self-citations are CAtNIPP [26] as architectural inspiration and [23] as a background citation for spatio-temporal attention; neither is load-bearing for the paper's validation, which relies on external baselines and the Gazebo simulation. No uniqueness theorem, no fitted parameter renamed as a prediction, and no ansatz smuggled in solely by self-citation were found. The derivation chain is therefore self-contained, with only minor, non-load-bearing self-citations, warranting a low circularity score.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The central claim depends on several fitted or hand-chosen parameters (GP hyperparameters, sensor model constants, ROI threshold, reward scale, graph structure) and on modeling assumptions (field is a GP, altitude noise model, PRM discretization, PPO convergence). These are standard in the field but are not independently validated against real sensor data in this paper.

free parameters (9)
  • GP length scale = 3.67
    Fit by maximizing log marginal likelihood over independent maps (Section V-A); sets the spatial correlation length of the belief model and strongly influences the reward landscape.
  • GP signal variance = 1.82
    Same fitting procedure; scales the covariance and hence the magnitude of uncertainty reduction rewards.
  • GP noise variance = 1.42
    Same fitting procedure; sets the sensor noise variance in the GP model used for belief updates.
  • ROI threshold µth = 0.4 in simulation; 0.05 in Gazebo
    Hand-chosen threshold to define regions of interest for reward and evaluation (Sections III-A, V-A, V-D).
  • Uncertainty weight β = not specified
    Design parameter in the ROI criterion µi + βσi ≥ µth; value not given in the text.
  • Sensor noise model parameters a and b = not specified
    User-chosen constants in σ² = a(1 - e^{-bh}) (Section III-D); no values provided, so the altitude-noise trade-off is not pinned down.
  • Reward scaling factor = 10
    Multiplier in Eq. (5) chosen to stabilize training by preventing vanishing gradients.
  • Number of graph neighbors k = 20
    Chosen for the PRM graph connectivity; the action set is the k nearest neighbors, so this directly shapes the policy's reachable states.
  • Altitude levels = 8 m and 14 m
    Two altitude levels used in all experiments; the core altitude-footprint trade-off depends on this discretization.
assumptions (5)
  • domain assumption The scalar field ζ is a Gaussian process with known kernel and hyperparameters.
    The belief representation, reward, and evaluation all use GP posterior updates; if the field is not GP-distributed, the learned policy is optimizing the wrong objective. Invoked in Sections III-B and V-A.
  • standard math The Kalman-like update (Eq. 3) correctly computes the GP posterior after measurements.
    Standard GP regression result, accepted as background.
  • ad hoc to paper Altitude-dependent sensor noise follows σ² = a(1 - e^{-bh}) and FoV expansion follows the described scaling.
    A chosen empirical model from prior work [9]; no validation against real sensor data in this paper. Drives the core altitude trade-off.
  • domain assumption The PRM graph with k nearest neighbors and k/n nodes per altitude adequately represents the continuous 3D planning space.
    The policy is restricted to graph nodes; a too-coarse graph would omit beneficial paths. Stated in Section IV-A.
  • domain assumption PPO training converges to a near-optimal policy for the constructed MDP.
    The paper relies on standard PPO convergence in practice; no convergence guarantees are given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Attention-based Learning for 3D Informative Path Planning." pith.science (2026). https://pith.science/paper/ZBWAPWGA

@misc{pith2026250608434,
  author       = {Pith},
  title        = {Pith review of: Attention-based Learning for 3D Informative Path Planning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZBWAPWGA}},
  note         = {Machine review of arXiv:2506.08434}
}
read the original abstract

In this work, we propose an attention-based deep reinforcement learning approach to address the adaptive informative path planning (IPP) problem in 3D space, where an aerial robot equipped with a downward-facing sensor must dynamically adjust its 3D position to balance sensing footprint and accuracy, and finally obtain a high-quality belief of an underlying field of interest over a given domain (e.g., presence of specific plants, hazardous gas, geological structures, etc.). In adaptive IPP tasks, the agent is tasked with maximizing information collected under time/distance constraints, continuously adapting its path based on newly acquired sensor data. To this end, we leverage attention mechanisms for their strong ability to capture global spatial dependencies across large action spaces, allowing the agent to learn an implicit estimation of environmental transitions. Our model builds a contextual belief representation over the entire domain, guiding sequential movement decisions that optimize both short- and long-term search objectives. Comparative evaluations against state-of-the-art planners demonstrate that our approach significantly reduces environmental uncertainty within constrained budgets, thus allowing the agent to effectively balance exploration and exploitation. We further show our model generalizes well to environments of varying sizes, highlighting its potential for many real-world applications.

Figures

Figures reproduced from arXiv: 2506.08434 by the authors.

Figure 1
Figure 1. Demonstration of our attention-based IPP planner in a weed-detection scenario. (Left) The UAV’s simulation environment, with scattered weed regions (green). (Right) A 3D path from blue to red over time, projecting weed occupancy (lighter cells) onto the grid. The high-density areas are adaptively prioritized in our work. However, recent progress in two-dimensional IPP [10], [11] does not fully leverage the ability o… view at source ↗
Figure 2
Figure 2. Sensor model and Kalman filter data fusion with agent belief. (a) Ground truth field distribution and the complete 3D probabilistic roadmap, where the agent’s current node (red) is connected to neighboring nodes (blue), with all other non-neighboring nodes shown in gray; (b) An independent measurement taken at 10m altitude, along with the corresponding belief update starting from an empty prior; (c) Another independ… view at source ↗
Figure 3
Figure 3. Our proposed attention-based neural network for 3D IPP. The encoder module employs a self-attention layer to capture global dependencies between nodes within the agent’s belief (i.e., the augmented graph) as context-aware node features. Leveraging both the current and neighboring context-aware node features (illustrated by the gray dashed circle), our decoder uses a “pointer network” structure with cross-attention, … view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Evaluation of our approach against baselines. On average, our RL approach ensures the fastest uncertainty (left) and RMSE (center left) reduction in regions of interest over mission time. Solid lines indicate means over 4 trials, and shaded regions associated standard …
Figure 6
Figure 6. Figure 6: Simulation environment and performance com￾parison. (Left) Top-down view of the simulated environment in Gazebo, where green regions indicate weed distributions. (Right) Trace of the covariance matrix tr(P) over mission time on 15 × 15 maps. Our method achieves faster …
Figure 5
Figure 5. Figure 5: Test results in 15 × 15 and 20 × 20 environments. Although our agent was only trained on 15 × 15 size maps, it outperforms all baselines on this larger map. D. Gazebo-based Simulation We demonstrate our approach through a Gazebo-based simulation in a weed detection sce…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

40 extracted references · 37 canonical work pages

  1. [27]

    Deep reinforcement learning with dynamic graphs for adaptive informative path planning

    Apoorva Vashisth, Julius Ruckin, Federico Magistri, Cyrill Stachniss, and Marija Popovic. Deep reinforcement learning with dynamic graphs for adaptive informative path planning. IEEE Robotics and Automation Letters, 9(9):7747–7754, 2024

    work page 2024

  2. [9]

    Adaptive informative path planning using deep reinforcement learning for uav-based active sensing

    Julius R ¨uckin, Liren Jin, and Marija Popovi ´c. Adaptive informative path planning using deep reinforcement learning for uav-based active sensing. In 2022 International Conference on Robotics and Automa- tion (ICRA), pages 4473–4479. IEEE, 2022

    work page 2022

  3. [1]

    Environmental reviews and case studies: bringing unmanned aerial systems closer to the environment

    Carrick Detweiler, John-Paul Ore, David Anthony, Sebastian Elbaum, Amy Burgin, and Aaron Lorenz. Environmental reviews and case studies: bringing unmanned aerial systems closer to the environment. Environmental Practice, 17(3):188–200, 2015

    work page 2015

  4. [2]

    Fully autonomous focused exploration for robotic environmental monitoring

    Gregory Hitz, Alkis Gotovos, Marie- ´Eve Garneau, C ´edric Pradalier, Andreas Krause, Roland Y Siegwart, et al. Fully autonomous focused exploration for robotic environmental monitoring. In 2014 IEEE International Conference on Robotics and Automation (ICRA) , pages 2658–2664. IEEE, 2014

    work page 2014

  5. [3]

    Robots for environmental monitoring: Significant advancements and applications

    Matthew Dunbabin and Lino Marques. Robots for environmental monitoring: Significant advancements and applications. IEEE Robotics & Automation Magazine , 19(1):24–39, 2012

    work page 2012

  6. [4]

    Autonomous gas-sensitive mi- crodrone: Wind vector estimation and gas distribution mapping

    Patrick P Neumann, Sahar Asadi, Achim J Lilienthal, Matthias Bartholmai, and Jochen H Schiller. Autonomous gas-sensitive mi- crodrone: Wind vector estimation and gas distribution mapping. IEEE robotics & automation magazine , 19(1):50–61, 2012

    work page 2012

  7. [5]

    Field testing of robotic technolo- gies to support ground ice prospecting in martian polygonal terrain

    Timothy D Barfoot, Paul T Furgale, Gordon R Osinski, Nadeem Ghafoor, and Kevin K Williams. Field testing of robotic technolo- gies to support ground ice prospecting in martian polygonal terrain. Planetary and Space Science , 58(4):671–681, 2010

    work page 2010

  8. [6]

    V olcanic environments: Robots for ex- ploration and measurement

    Giovanni Muscato, Filippo Bonaccorso, Luciano Cantelli, Domenico Longo, and C Donato Melita. V olcanic environments: Robots for ex- ploration and measurement. IEEE Robotics & Automation Magazine , 19(1):40–49, 2012

    work page 2012

Show all 40 references
  1. [7]

    Adaptive continuous-space informative path planning for online environmental monitoring

    Gregory Hitz, Enric Galceran, Marie- `Eve Garneau, Franc ¸ois Pomer- leau, and Roland Siegwart. Adaptive continuous-space informative path planning for online environmental monitoring. Journal of Field Robotics, 34(8):1427–1449, 2017

    2017

  2. [8]

    Online informative path planning for active classification using uavs

    Marija Popovi ´c, Gregory Hitz, Juan Nieto, Inkyu Sa, Roland Siegwart, and Enric Galceran. Online informative path planning for active classification using uavs. In 2017 IEEE international conference on robotics and automation (ICRA) , pages 5753–5758. IEEE, 2017

    2017

  3. [10]

    Informative path planning for active field mapping under localization uncertainty

    Marija Popovi ´c, Teresa Vidal-Calleja, Jen Jen Chung, Juan Nieto, and Roland Siegwart. Informative path planning for active field mapping under localization uncertainty. In 2020 IEEE International Conference on Robotics and Automation (ICRA), pages 10751–10757. IEEE, 2020

    2020

  4. [11]

    Adaptive informative path planning with multimodal sensing

    Shushman Choudhury, Nate Gruver, and Mykel J Kochenderfer. Adaptive informative path planning with multimodal sensing. In Proceedings of the International Conference on Automated Planning and Scheduling, volume 30, pages 57–65, 2020

    2020

  5. [12]

    Autonomous exploration under uncertainty via deep reinforce- ment learning on graphs

    Fanfei Chen, John D Martin, Yewei Huang, Jinkun Wang, and Brendan Englot. Autonomous exploration under uncertainty via deep reinforce- ment learning on graphs. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pages 6140–6147. IEEE, 2020

    2020

  6. [13]

    Deepig: Multi-robot information gathering with deep reinforcement learning

    Alberto Viseras and Ricardo Garcia. Deepig: Multi-robot information gathering with deep reinforcement learning. IEEE Robotics and Automation Letters, 4(3):3059–3066, 2019

    2019

  7. [14]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems , 30, 2017

    2017

  8. [15]

    Planning and acting in partially observable stochastic domains

    Leslie Pack Kaelbling, Michael L Littman, and Anthony R Cassan- dra. Planning and acting in partially observable stochastic domains. Artificial intelligence, 101(1-2):99–134, 1998

    1998

  9. [16]

    Information-theoretic planning with trajectory optimization for dense 3d mapping

    Benjamin Charrow, Gregory Kahn, Sachin Patil, Sikang Liu, Ken Goldberg, Pieter Abbeel, Nathan Michael, and Vijay Kumar. Information-theoretic planning with trajectory optimization for dense 3d mapping. In Robotics: Science and Systems , volume 11, pages 3–12. Rome, 2015

    2015

  10. [17]

    Sampling-based robotic information gathering algorithms

    Geoffrey A Hollinger and Gaurav S Sukhatme. Sampling-based robotic information gathering algorithms. The International Journal of Robotics Research , 33(9):1271–1287, 2014

    2014

  11. [18]

    Gaussian process oc- cupancy maps

    Simon T O’Callaghan and Fabio T Ramos. Gaussian process oc- cupancy maps. The International Journal of Robotics Research , 31(1):42–62, 2012

    2012

  12. [19]

    Bayesian fusion using conditionally independent submaps for high resolution 2.5 d mapping

    Liye Sun, Teresa Vidal-Calleja, and Jaime Valls Miro. Bayesian fusion using conditionally independent submaps for high resolution 2.5 d mapping. In 2015 IEEE International Conference on Robotics and Automation (ICRA), pages 3394–3400. IEEE, 2015

    2015

  13. [20]

    Gaussian process modeling of large-scale terrain

    Shrihari Vasudevan, Fabio Ramos, Eric Nettleton, and Hugh Durrant- Whyte. Gaussian process modeling of large-scale terrain. Journal of Field Robotics, 26(10):812–840, 2009

    2009

  14. [21]

    Op- timizing waypoints for monitoring spatiotemporal phenomena

    Jonathan Binney, Andreas Krause, and Gaurav S Sukhatme. Op- timizing waypoints for monitoring spatiotemporal phenomena. The International Journal of Robotics Research , 32(8):873–888, 2013

    2013

  15. [22]

    Gaussian processes for machine learning

    Matthias Seeger. Gaussian processes for machine learning. Interna- tional journal of neural systems , 14(02):69–106, 2004

    2004

  16. [23]

    Spatio-temporal attention network for persistent monitoring of mul- tiple mobile targets

    Yizhuo Wang, Yutong Wang, Yuhong Cao, and Guillaume Sartoretti. Spatio-temporal attention network for persistent monitoring of mul- tiple mobile targets. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 3903–3910. IEEE, 2023

    2023

  17. [24]

    Modeling and decision making in spatio-temporal processes for environmental surveillance

    Amarjeet Singh, Fabio Ramos, Hugh Durrant Whyte, and William J Kaiser. Modeling and decision making in spatio-temporal processes for environmental surveillance. In 2010 IEEE International Conference on Robotics and Automation , pages 5490–5497. IEEE, 2010

    2010

  18. [25]

    Randomized algorithm for informative path planning with budget constraints

    Sankalp Arora and Sebastian Scherer. Randomized algorithm for informative path planning with budget constraints. In 2017 IEEE International Conference on Robotics and Automation (ICRA) , pages 4997–5004. IEEE, 2017

    2017

  19. [26]

    Catnipp: Context-aware attention-based network for informative path planning

    Yuhong Cao, Yizhuo Wang, Apoorva Vashisth, Haolin Fan, and Guillaume Adrien Sartoretti. Catnipp: Context-aware attention-based network for informative path planning. In Conference on Robot Learning, pages 1928–1937. PMLR, 2023

    1928

  20. [28]

    Near-optimal sensor placements in gaussian processes: Theory, efficient algorithms and empirical studies

    Andreas Krause, Ajit Singh, and Carlos Guestrin. Near-optimal sensor placements in gaussian processes: Theory, efficient algorithms and empirical studies. Journal of Machine Learning Research , 9(2), 2008

    2008

  21. [29]

    Complexity of the mover’s problem and generalizations

    John H Reif. Complexity of the mover’s problem and generalizations. In 20th Annual Symposium on Foundations of Computer Science (sfcs 1979), pages 421–427. IEEE Computer Society, 1979

    1979

  22. [30]

    A recursive greedy algorithm for walks in directed graphs

    Chandra Chekuri and Martin Pal. A recursive greedy algorithm for walks in directed graphs. In 46th annual IEEE symposium on foundations of computer science (FOCS’05) , pages 245–253. IEEE, 2005

    2005

  23. [31]

    Efficient multi-robot search for a moving target

    Geoffrey Hollinger, Sanjiv Singh, Joseph Djugash, and Athanasios Kehagias. Efficient multi-robot search for a moving target. The International Journal of Robotics Research , 28(2):201–219, 2009

    2009

  24. [32]

    Branch and bound for informative path planning

    Jonathan Binney and Gaurav S Sukhatme. Branch and bound for informative path planning. In 2012 IEEE international conference on robotics and automation , pages 2147–2154. IEEE, 2012

    2012

  25. [33]

    An introduction to gaussian processes for the kalman filter expert

    Steven Reece and Stephen Roberts. An introduction to gaussian processes for the kalman filter expert. In 2010 13th International Conference on Information Fusion , pages 1–9. IEEE, 2010

    2010

  26. [34]

    Adaptive-resolution field mapping using gaussian process fusion with integral kernels

    Liren Jin, Julius R ¨uckin, Stefan H Kiss, Teresa Vidal-Calleja, and Marija Popovi ´c. Adaptive-resolution field mapping using gaussian process fusion with integral kernels. IEEE Robotics and Automation Letters, 7(3):7471–7478, 2022

    2022

  27. [35]

    A comparative study of probabilistic roadmap planners

    Roland Geraerts and Mark H Overmars. A comparative study of probabilistic roadmap planners. Algorithmic foundations of robotics V, pages 43–57, 2004

    2004

  28. [36]

    Benchmarking graph neural networks

    Vijay Prakash Dwivedi, Chaitanya K Joshi, Thomas Laurent, Yoshua Bengio, and Xavier Bresson. Benchmarking graph neural networks. 2020

    2020

  29. [37]

    Pointer networks

    Oriol Vinyals, Meire Fortunato, and Navdeep Jaitly. Pointer networks. Advances in neural information processing systems , 28, 2015

    2015

  30. [38]

    Proximal policy optimization algorithms

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017

    2017 arXiv

  31. [39]

    Using occupancy grids for mobile robot perception and navigation

    Alberto Elfes. Using occupancy grids for mobile robot perception and navigation. Computer, 22(6):46–57, 1989

    1989

  32. [40]

    Online algorithms for pomdps with continuous state, action, and observation spaces

    Zachary Sunberg and Mykel Kochenderfer. Online algorithms for pomdps with continuous state, action, and observation spaces. In Proceedings of the International Conference on Automated Planning and Scheduling, volume 28, pages 259–263, 2018

    2018

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.