Pith. sign in

REVIEW 3 major objections 7 minor 39 references

What Matters in Learning A Zero-Shot Sim-to-Real RL Policy for Quadrotor Control? A Comprehensive Study

T0 review · 3 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read SimpleFlight claims that five concrete training choices—velocity and rotation-matrix actor inputs, critic time vector, action-difference smoothness, selective domain randomization, and large batches—make a PPO quadrotor policy transfer…

desk verdict Useful empirical recipe for zero-shot quadrotor RL, but the headline margin is partly in-sample because key hyperparameters were tuned on the same real-world tasks used for evaluation. read the letter →

arxiv 2412.11764 v4 pith:6OXUH6ZR submitted 2024-12-16 cs.RO cs.LG

Jiayu Chen , Chao Yu , Yuqing Xie , Feng Gao , Yinuo Chen , Shu'ang Yu , Wenhao Tang , Shilong Ji
show 4 more authors
Mo Mu Yi Wu Huazhong Yang Yu Wang
This is my paper · ORCID
classification cs.ROcs.LG
keywords zero-shotsim-to-realquadrotorcontrolreinforcementlearningtrajectorytrackingdomainrandomizationsystemidentificationsmoothnessregularizationproximalpolicyoptimization
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that zero-shot sim-to-real transfer for quadrotor control does not require a new algorithm or a bespoke network architecture: five concrete training choices are enough. The choices are feeding the actor velocity and a rotation matrix rather than a quaternion, giving the critic a time vector, penalizing the difference between successive actions, calibrating key dynamic parameters with domain randomization only where it helps, and using large batch sizes. The authors package these into a PPO-based framework called SimpleFlight and report that on a nano quadrotor it reduces trajectory tracking error by more than 50% relative to state-of-the-art RL baselines. Because the same policy, deployed without fine-tuning, also completes infeasible zigzag trajectories that baselines fail, the recipe claims to cover both smooth and aggressive flight.

What carries the argument

The machinery is the SimpleFlight training pipeline: an asymmetric actor-critic PPO setup whose actions are CTBR commands (collective thrust and body rates). The actor observes relative positions to ten future reference points, linear velocity, and the rotation matrix; the critic observes the same plus a scalar time vector. The reward augments the tracking task with $e^{-\|u_t-u_{t-1}\|_2}$, which directly penalizes jumps between successive commands. Before training, four dynamic parameters (mass, inertia, thrust coefficient, and motor time constant) are calibrated, and domain randomization is applied selectively, principally to the thrust coefficient; the low-level controller inside the simulator is aligned to the firmware's response. Training runs in a GPU-parallel simulator with large batches, and the resulting policy is deployed without modification.

What would settle it

Run the paper's Offset+30% condition on a real quadrotor: if deliberately shifting the thrust coefficient $k_f$ by +30% does not raise tracking error well above the calibrated policy's error, the claim that $k_f$ is a sensitive parameter and that system identification is load-bearing is contradicted. A complementary test is to compare simulated and real motor step responses, since the low-level controller alignment premise must hold for any zero-shot transfer.

Watch

Extended reading notes

Core claim

SimpleFlight claims that a PPO-based policy trained with five specific choices—velocity and rotation matrix in the actor's observation, a time vector in the critic's observation, an action-difference smoothness reward $\|u_t-u_{t-1}\|_2$, calibrated dynamics with domain randomization applied only to the thrust coefficient, and large training batch sizes—can be deployed directly on real quadrotors without fine-tuning. In real-world tests on a nano quadrotor, it reports a reduction of more than 50% in mean Euclidean distance to the reference trajectory compared with state-of-the-art RL baselines, and it is the only tested policy that completes all benchmark trajectories, including fast pentagram and infeasible zigzag paths. The same policy class also transfers to a larger custom quadrotor, where it slightly outperforms a finely tuned model-predictive controller. The paper's claim is that the five factors, not any new algorithm or architecture, are what make zero-shot transfer work.

Load-bearing premise

The whole zero-shot claim rests on the simulated low-level controller and the rigid-body dynamics matching the real quadrotor's firmware and airframe closely enough that a policy trained in simulation faces the same input-output mapping on hardware.

Editorial extensions

If this is right

  • Any future RL quadrotor controller that adopts these five factors should expect at least a halving of trajectory tracking error on smooth trajectories compared with current RL baselines, under zero-shot deployment.
  • Small thrust-to-weight quadrotors can track infeasible zigzag trajectories that previously required optimization-based methods, because the action-difference smoothness regularizer supports long-horizon planning without clipping agility.
  • Larger batch sizes are a free lunch: they improve real-world generalization with no change to architecture or reward.
  • Domain randomization should be applied selectively; randomizing well-calibrated mass or inertia can hurt real-world performance, while randomizing the thrust coefficient helps when calibration is imperfect.
  • The recipe is platform-transferable: the same policy class transfers to a second, larger custom quadrotor and remains competitive with a finely tuned MPC controller.

Reading between the lines

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

  • The paper's ablations imply that rotation-matrix inputs are a first-order factor for agile flight, so the same representation choice likely matters for other rotation-controlled robots, not just quadrotors.
  • Since the lookahead horizon is fixed at $N=10$ without ablation, a natural extension is to sweep $N$ on infeasible trajectories, where longer horizons should matter most.
  • The selective-domain-randomization result suggests a practical tuning rule: randomize only parameters that are hard to measure and performance-sensitive, and leave precisely measurable ones fixed.
  • The time-vector result hints that adding task-phase information to value networks could be a cheap general trick for non-stationary RL problems beyond quadrotor control.
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

3 major / 7 minor

Summary. The paper identifies five factors for learning zero-shot sim-to-real RL policies for quadrotor trajectory tracking: actor inputs including velocity and rotation matrix, critic-only time vector, action-difference smoothness regularization, system identification with selective domain randomization, and large batch sizes. These factors are combined into a PPO-based framework called SimpleFlight. The authors validate SimpleFlight on a Crazyflie 2.1 and a custom 'Air' quadrotor, reporting over 50% reduction in mean Euclidean distance tracking error against two RL baselines (Fly, DATT) and comparable or better performance than a fine-tuned MPC baseline (PAMPC). The paper also integrates SimpleFlight into the Omnidrones simulator and open-sources code and checkpoints.

Significance. If the central claim holds, this is a valuable empirical study that consolidates several known-but-scattered design choices into a concrete recipe and demonstrates real-world viability on low-thrust-to-weight quadrotors. The open-source release and Omnidrones integration are concrete assets for reproducibility. However, the headline quantitative claim is compromised by the fact that several recipe components were selected using real-world performance on the same figure-eight benchmark that appears in the final evaluation, and the paper itself acknowledges that the baseline comparison is not entirely fair. These issues make the reported 50% improvement an in-sample estimate rather than a validated zero-shot prediction, so the strongest claims in the abstract and Section V need revision or additional holdout experiments before they can be accepted as stated.

major comments (3)
  1. [§V-B2, V-B3, V-B4; Fig. 4; Table II; Fig. 5; Table III] The smoothness reward coefficient λ is set to 0.4 based on real-world figure-eight tracking performance (Fig. 4), the domain-randomization ranges and conclusions are derived from real-world figure-eight experiments (Table II), and the batch-size recommendation is based on real-world figure-eight curves (Fig. 5). These same figure-eight trajectories (slow, normal, fast) are then reported in the final comparison of Table III. Consequently, the reported "more than 50% reduction" and the "only one capable of successfully completing all benchmarking trajectories" claims are in-sample estimates rather than predictive evidence for a zero-shot recipe. The paper should provide a holdout evaluation on trajectory types or parameter settings that were not used for any hyperparameter selection, or it should explicitly state that the final comparison shares the tuning tasks and correspondingly soften the zero-shot claim.
  2. [§V-C2 and Table III] The paper states that the comparison in Table III "may not be entirely fair" because the policies use different simulators, modeling approaches, and input/output spaces. Since the headline claim of over 50% error reduction is relative to these baselines, this caveat should be stated in the abstract and in the main text where the claim is made, not only at the end of Section V-C2. Additionally, for DATT on zigzag trajectories, the footnote reports that 4 out of 10 attempts failed and only the successful trials are averaged; the failure rate itself is a primary metric and should be reported in the table or text before any claim that SimpleFlight is "the only one" capable of completing all benchmarks.
  3. [§V-B3 and Table II] The selective-DR conclusion is based on artificially induced +30% parameter offsets to simulate miscalibration. While this is a reasonable synthetic test, the real calibration errors of the Crazyflie are not quantified, so the practical recommendation that DR should be applied only to thrust coefficient kf is not directly supported by measured calibration uncertainty. A sensitivity analysis over the actual estimated parameter errors, or a statement of the calibration accuracy, would strengthen the factor-4 claim.
minor comments (7)
  1. [Abstract] The word "polynominal" is a typo and should be "polynomial."
  2. [Section I] The text "U A Vs" in the introduction should be "UAVs" without spaces.
  3. [Section III-B] The text "17-dimensinal" is a typo and should be "17-dimensional."
  4. [Section V-B1] The observed "approximately 63.6%" performance degradation when replacing the rotation matrix with a quaternion is not derived from any displayed curve; please specify the exact quantities or figure from which this percentage is computed.
  5. [Section V-B2 and Table I] The smoothness reward is presented in Eq. (4) as a generic rsmooth, but Table I evaluates components using raux = e^{-A}; please clarify the relationship between these two formulations, particularly how ||ut - ut-1||2 is inserted as A.
  6. [Table III] For trajectories where some trials fail, the MED is reported only over successful trials (e.g., DATT zigzag). Please add a column or annotation showing the number of successful trials per method and trajectory, since "∞" alone does not distinguish partial from total failure.
  7. [References] References [7] and [39] both cite Williams et al. on information-theoretic MPC; please merge or differentiate them.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the five-factor recipe is an empirical ablation validated by real-world deployment; the figure-eight hyperparameter selection is a test-set fairness caveat, not a circular derivation.

full rationale

The central claim is an experimental result: PPO policies trained in OmniDrones with five specified factors are deployed zero-shot on Crazyflie and Air, and their measured mean Euclidean distances are compared with baselines. The five factors are identified through ablations (Fig. 3, Tab. I, Tab. II, Fig. 5), not derived from the final metric by construction. The reported tracking errors are determined by simulation training and real flight measurements, not by a fitted equation, so no step equates the prediction to its input by construction. The self-citation to OmniDrones [17] is a tool citation and is not load-bearing: the transfer results are real-world measurements, and the simulator is open-source. The only in-scope caveat is that lambda, DR ranges, and batch size are selected using real-world figure-eight experiments (Sec. V-B2: 'Based on these findings, we set lambda = 0.4'; Sec. V-B3; Sec. V-B4: 'we test simulation and real-world performance using figure-eight trajectories'), and the same figure-eight task reappears in the final comparison in Tab. III. This makes the figure-eight column an in-sample estimate rather than an independent prediction, and the paper's own remark that the Tab. III comparison 'may not be entirely fair' reinforces caution about the exact margins. However, this is test-set selection, not circular derivation: the figure-eight MED is not equal to lambda or to the tuning objective by construction, and the central zero-shot claim also rests on out-of-distribution pentagram/zigzag results and cross-platform Air experiments that were not tuned on those real-world tasks. No self-definitional, uniqueness-import, ansatz-smuggling, or renaming circularity is present.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities. The free parameters are either hyperparameters tuned on real-world data (lambda, N, k) or platform-specific physical parameters calibrated via system identification. The main axioms are the fidelity of the simulator model and the alignment of the low-level controller, both of which are load-bearing for zero-shot transfer.

free parameters (7)
  • lambda (smoothness reward weight) = 0.4
    Selected by grid search on real-world figure-eight tracking error (Fig. 4); appears in reward Eq. (4) and directly affects the trade-off between task reward and smoothness.
  • N (lookahead waypoint count) = 10
    Chosen by hand; each waypoint spaced 0.05s; defines the actor's long-horizon planning input.
  • k (time vector dimension) = 1
    Set to 1 because the authors found training performance insensitive to k (Section IV-B); included in the critic's privileged input.
  • Mass m = not reported
    Calibrated via SysID; used in simulator dynamics Eq. (1).
  • Inertia matrix I = not reported
    Calibrated via SysID; used in Eq. (1).
  • Thrust coefficient k_f = not reported
    Calibrated via static propeller test stand; selectively randomized in the final policy.
  • Motor time constant T_m = not reported
    Calibrated via SysID; used in first-order motor model.
assumptions (5)
  • domain assumption Quadrotor dynamics follow a 6-DOF rigid-body model with first-order motor response (Eq. (1)-(3)).
    The entire simulation and transfer rely on this model; unmodeled aerodynamic drag and motor nonlinearities are ignored.
  • domain assumption After calibration, the simulated low-level CTBR controller matches the real firmware.
    Central to zero-shot transfer; stated in Section IV-A but with limited detail.
  • standard math PPO with the specified MLP architecture can represent the optimal control policy.
    The paper relies on PPO's convergence and the universal approximation capacity of the 3-layer MLP, which is standard in RL practice.
  • domain assumption Motion capture provides accurate state at 100 Hz and the offboard control loop meets the timing.
    Real-world deployment protocol uses OptiTrack and offboard computation; timing or state errors would affect results.
  • ad hoc to paper The five selected factors are the critical determinants of zero-shot sim-to-real performance.
    This is the paper's core thesis, but it is based on ablations of a fixed set of factors; other unexplored factors (e.g., observation noise, reward shaping details, controller frequency) are not evaluated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of What Matters in Learning A Zero-Shot Sim-to-Real RL Policy for Quadrotor Control? A Comprehensive Study." pith.science (2026). https://pith.science/paper/6OXUH6ZR

@misc{pith2026241211764,
  author       = {Pith},
  title        = {Pith review of: What Matters in Learning A Zero-Shot Sim-to-Real RL Policy for Quadrotor Control? A Comprehensive Study},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6OXUH6ZR}},
  note         = {Machine review of arXiv:2412.11764}
}
read the original abstract

Executing precise and agile flight maneuvers is critical for quadrotors in various applications. Traditional quadrotor control approaches are limited by their reliance on flat trajectories or time-consuming optimization, which restricts their flexibility. Recently, RL-based policy has emerged as a promising alternative due to its ability to directly map observations to actions, reducing the need for detailed system knowledge and actuation constraints. However, a significant challenge remains in bridging the sim-to-real gap, where RL-based policies often experience instability when deployed in real world. In this paper, we investigate key factors for learning robust RL-based control policies that are capable of zero-shot deployment in real-world quadrotors. We identify five critical factors and we develop a PPO-based training framework named SimpleFlight, which integrates these five techniques. We validate the efficacy of SimpleFlight on Crazyflie quadrotor, demonstrating that it achieves more than a 50% reduction in trajectory tracking error compared to state-of-the-art RL baselines. The policy derived by SimpleFlight consistently excels across both smooth polynominal trajectories and challenging infeasible zigzag trajectories on small thrust-to-weight quadrotors. In contrast, baseline methods struggle with high-speed or infeasible trajectories. To support further research and reproducibility, we integrate SimpleFlight into a GPU-based simulator Omnidrones and provide open-source access to the code and model checkpoints. We hope SimpleFlight will offer valuable insights for advancing RL-based quadrotor control. For more details, visit our project website at https://sites.google.com/view/simpleflight/.

Figures

Figures reproduced from arXiv: 2412.11764 by the authors.

Figure 1
Figure 1. Overview of SimpleFlight. We begin with SysID and selective DR for quadrotor dynamics and low-level control. Next, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Visualization of benchmark trajectories and corresponding trajectories followed using SimpleFlight. The reference [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Training performance of input space designs. (a) For [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Real-world performance of different λ on the figure￾eight trajectory. We finally choose λ = 0.4. (a) Slow. (b) Normal. (c) Fast [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 25 canonical work pages

  1. [1]

    Low-cost autonomous uav- based solutions to package delivery logistics,

    J. Grzybowski, K. Latos, and R. Czyba, “Low-cost autonomous uav- based solutions to package delivery logistics,” in Advanced, Contemporary Control: Proceedings of KKA 2020—The 20th Polish Control Conference, Łód´ z, Poland, 2020. Springer, 2020, pp. 500–507

    work page 2020

  2. [2]

    An autonomous multi-uav system for search and rescue,

    J. Scherer, S. Yahyanejad, S. Hayat, E. Yanmaz, T. Andre, A. Khan, V . Vukadinovic, C. Bettstetter, H. Hellwagner, and B. Rinner, “An autonomous multi-uav system for search and rescue,” in Proceedings of the first workshop on micro aerial vehicle networks, systems, and applications for civilian use , 2015, pp. 33–38

    work page 2015

  3. [3]

    A uav system for inspection of industrial facilities,

    J. Nikolic, M. Burri, J. Rehder, S. Leutenegger, C. Huerzeler, and R. Siegwart, “A uav system for inspection of industrial facilities,” in 2013 IEEE Aerospace Conference . IEEE, 2013, pp. 1–8

    work page 2013

  4. [4]

    Minimum snap trajectory generation and control for quadrotors,

    D. Mellinger and V . Kumar, “Minimum snap trajectory generation and control for quadrotors,” in 2011 IEEE International Conference on Robotics and Automation , 2011, pp. 2520–2525

    2011

  5. [5]

    Differential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high- speed trajectories,

    M. Faessler, A. Franchi, and D. Scaramuzza, “Differential flatness of quadrotor dynamics subject to rotor drag for accurate tracking of high- speed trajectories,” IEEE Robotics and Automation Letters , vol. 3, no. 2, pp. 620–626, 2018

    work page 2018

  6. [6]

    Performance, precision, and payloads: Adaptive nonlinear mpc for quadrotors,

    D. Hanover, P. Foehn, S. Sun, E. Kaufmann, and D. Scaramuzza, “Performance, precision, and payloads: Adaptive nonlinear mpc for quadrotors,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 690–697, 2021

    work page 2021

  7. [7]

    Information theoretic mpc for model-based reinforcement learning,

    G. Williams, N. Wagener, B. Goldfain, P. Drews, J. M. Rehg, B. Boots, and E. A. Theodorou, “Information theoretic mpc for model-based reinforcement learning,” in 2017 IEEE International Conference on Robotics and Automation (ICRA) , 2017, pp. 1714–1721

    2017

  8. [8]

    Control of a quadrotor with reinforcement learning,

    J. Hwangbo, I. Sa, R. Siegwart, and M. Hutter, “Control of a quadrotor with reinforcement learning,” IEEE Robotics and Automation Letters , vol. 2, no. 4, pp. 2096–2103, 2017

    2017

Show all 39 references
  1. [9]

    Optimal and autonomous control using reinforcement learning: A survey,

    B. Kiumarsi, K. G. Vamvoudakis, H. Modares, and F. L. Lewis, “Optimal and autonomous control using reinforcement learning: A survey,” IEEE transactions on neural networks and learning systems , vol. 29, no. 6, pp. 2042–2062, 2017

    2017

  2. [10]

    Visual attention prediction improves performance of autonomous drone racing agents,

    C. Pfeiffer, S. Wengeler, A. Loquercio, and D. Scaramuzza, “Visual attention prediction improves performance of autonomous drone racing agents,” Plos one, vol. 17, no. 3, p. e0264471, 2022

    2022

  3. [11]

    Datt: Deep adaptive trajectory tracking for quadrotor control,

    K. Huang, R. Rana, A. Spitzer, G. Shi, and B. Boots, “Datt: Deep adaptive trajectory tracking for quadrotor control,” in Conference on Robot Learning. PMLR, 2023, pp. 326–340

    2023

  4. [12]

    Neural internal model control: Learning a robust control policy via predictive error feedback,

    F. Gao, C. Yu, Y . Wang, and Y . Wu, “Neural internal model control: Learning a robust control policy via predictive error feedback,” arXiv preprint arXiv: 2411.13079 , 2024

    2024 arXiv

  5. [13]

    Autonomous drone racing with deep reinforcement learning,

    Y . Song, M. Steinweg, E. Kaufmann, and D. Scaramuzza, “Autonomous drone racing with deep reinforcement learning,” in 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2021, pp. 1205–1212

    2021

  6. [14]

    Learning to fly in seconds,

    J. Eschmann, D. Albani, and G. Loianno, “Learning to fly in seconds,” IEEE Robotics and Automation Letters , vol. 9, no. 7, pp. 6336–6343, 2024

    2024

  7. [15]

    A benchmark comparison of learned control policies for agile quadrotor flight,

    E. Kaufmann, L. Bauersfeld, and D. Scaramuzza, “A benchmark comparison of learned control policies for agile quadrotor flight,” in 2022 International Conference on Robotics and Automation (ICRA) . IEEE, 2022, pp. 10 504–10 510

    2022

  8. [16]

    The power of input: Benchmarking zero-shot sim-to-real transfer of reinforcement learning control policies for quadrotor control,

    A. Dionigi, G. Costante, and G. Loianno, “The power of input: Benchmarking zero-shot sim-to-real transfer of reinforcement learning control policies for quadrotor control,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2024, pp. 11 ...

    2024

  9. [17]

    Omnidrones: An efficient and flexible platform for reinforcement learning in drone control,

    B. Xu, F. Gao, C. Yu, R. Zhang, Y . Wu, and Y . Wang, “Omnidrones: An efficient and flexible platform for reinforcement learning in drone control,” IEEE Robotics and Automation Letters , 2024

    2024

  10. [18]

    System identification and control using genetic algorithms,

    K. Kristinsson and G. A. Dumont, “System identification and control using genetic algorithms,” IEEE Transactions on Systems, Man, and Cybernetics, vol. 22, no. 5, pp. 1033–1046, 1992

    1992

  11. [19]

    Domain randomization for transferring deep neural networks from simulation to the real world,

    J. Tobin, R. Fong, A. Ray, J. Schneider, W. Zaremba, and P. Abbeel, “Domain randomization for transferring deep neural networks from simulation to the real world,” in 2017 IEEE/RSJ international conference on intelligent robots and systems (IROS) . IEEE, 2017, pp. 23–30

    2017

  12. [20]

    Training deep networks with synthetic data: Bridging the reality gap by domain randomization,

    J. Tremblay, A. Prakash, D. Acuna, M. Brophy, V . Jampani, C. Anil, T. To, E. Cameracci, S. Boochoon, and S. Birchfield, “Training deep networks with synthetic data: Bridging the reality gap by domain randomization,” in Proceedings of the IEEE conference on computer vision and...

    2018

  13. [21]

    Learning dexterous in-hand manipulation,

    O. M. Andrychowicz, B. Baker, M. Chociej, R. Jozefowicz, B. McGrew, J. Pachocki, A. Petron, M. Plappert, G. Powell, A. Ray et al., “Learning dexterous in-hand manipulation,” The International Journal of Robotics Research, vol. 39, no. 1, pp. 3–20, 2020

    2020

  14. [22]

    Sim-to-(multi)-real: Transfer of low-level robust control policies to multiple quadrotors,

    A. Molchanov, T. Chen, W. Hönig, J. A. Preiss, N. Ayanian, and G. S. Sukhatme, “Sim-to-(multi)-real: Transfer of low-level robust control policies to multiple quadrotors,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2019, pp. 59–66

    2019

  15. [23]

    Deep drone racing: From simulation to reality with domain randomization,

    A. Loquercio, E. Kaufmann, R. Ranftl, A. Dosovitskiy, V . Koltun, and D. Scaramuzza, “Deep drone racing: From simulation to reality with domain randomization,” IEEE Transactions on Robotics , vol. 36, no. 1, pp. 1–14, 2019

    2019

  16. [24]

    Domain-adversarial training of neural networks,

    Y . Ganin, E. Ustinova, H. Ajakan, P. Germain, H. Larochelle, F. Laviolette, M. March, and V . Lempitsky, “Domain-adversarial training of neural networks,” Journal of machine learning research , vol. 17, no. 59, pp. 1–35, 2016

    2016

  17. [25]

    Unsupervised pixel-level domain adaptation with generative adversarial networks,

    K. Bousmalis, N. Silberman, D. Dohan, D. Erhan, and D. Krishnan, “Unsupervised pixel-level domain adaptation with generative adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 3722–3731

    2017

  18. [26]

    Unsupervised adversarial domain adaptation for sim-to-real transfer of tactile images,

    X. Jing, K. Qian, T. Jianu, and S. Luo, “Unsupervised adversarial domain adaptation for sim-to-real transfer of tactile images,” IEEE Transactions on Instrumentation and Measurement , vol. 72, pp. 1–11, 2023

    2023

  19. [27]

    Learning transferable features with deep adaptation networks,

    M. Long, Y . Cao, J. Wang, and M. Jordan, “Learning transferable features with deep adaptation networks,” in International conference on machine learning. PMLR, 2015, pp. 97–105

    2015

  20. [28]

    Sim-to-real visual grasping via state representation learning based on combining pixel-level and feature-level domain adaptation,

    Y . Park, S. H. Lee, and I. H. Suh, “Sim-to-real visual grasping via state representation learning based on combining pixel-level and feature-level domain adaptation,” in 2021 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2021, pp. 6300–6307

    2021

  21. [29]

    Using simulation optimization to improve zero-shot policy transfer of quadrotors,

    S. Gronauer, M. Kissel, L. Sacchetto, M. Korte, and K. Diepold, “Using simulation optimization to improve zero-shot policy transfer of quadrotors,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2022, pp. 10 170–10 176

    2022

  22. [30]

    Neurobem: Hybrid aerodynamic quadrotor model,

    L. Bauersfeld, E. Kaufmann, P. Foehn, S. Sun, and D. Scaramuzza, “Neurobem: Hybrid aerodynamic quadrotor model,” arXiv preprint arXiv:2106.08015, 2021

    2021 arXiv

  23. [31]

    Champion-level drone racing using deep reinforcement learning,

    E. Kaufmann, L. Bauersfeld, A. Loquercio, M. Müller, V . Koltun, and D. Scaramuzza, “Champion-level drone racing using deep reinforcement learning,” Nature, vol. 620, no. 7976, pp. 982–987, 2023

    2023

  24. [32]

    Learning a single near-hover position controller for vastly different quadcopters,

    D. Zhang, A. Loquercio, X. Wu, A. Kumar, J. Malik, and M. W. Mueller, “Learning a single near-hover position controller for vastly different quadcopters,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 1263–1269

    2023

  25. [33]

    Learning to fly in seconds,

    J. Eschmann, D. Albani, and G. Loianno, “Learning to fly in seconds,” IEEE Robotics and Automation Letters , 2024

    2024

  26. [34]

    Rotors—a modular gazebo mav simulator framework,

    F. Furrer, M. Burri, M. Achtelik, and R. Siegwart, “Rotors—a modular gazebo mav simulator framework,” Robot Operating System (ROS) The Complete Reference (Volume 1) , pp. 595–625, 2016

    2016

  27. [35]

    Airsim: High-fidelity visual and physical simulation for autonomous vehicles,

    S. Shah, D. Dey, C. Lovett, and A. Kapoor, “Airsim: High-fidelity visual and physical simulation for autonomous vehicles,” in Field and Service Robotics: Results of the 11th International Conference . Springer, 2018, pp. 621–635

    2018

  28. [36]

    Proxi- mal policy optimization algorithms,

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proxi- mal policy optimization algorithms,” arXiv preprint arXiv:1707.06347 , 2017

    2017 arXiv

  29. [37]

    On the continuity of rotation representations in neural networks,

    Y . Zhou, C. Barnes, J. Lu, J. Yang, and H. Li, “On the continuity of rotation representations in neural networks,” in 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2019, pp. 5738–5746

    2019

  30. [38]

    Pampc: Perception- aware model predictive control for quadrotors,

    D. Falanga, P. Foehn, P. Lu, and D. Scaramuzza, “Pampc: Perception- aware model predictive control for quadrotors,” in 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2018, pp. 1–8

    2018

  31. [39]

    Information theoretic mpc for model-based reinforcement learning,

    G. Williams, N. Wagener, B. Goldfain, P. Drews, J. M. Rehg, B. Boots, and E. A. Theodorou, “Information theoretic mpc for model-based reinforcement learning,” in 2017 IEEE international conference on robotics and automation (ICRA) . IEEE, 2017, pp. 1714–1721

    2017

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.