Pith. sign in

REVIEW 3 major objections 5 minor 42 references

Toward Sim-to-Real Directional Semantic Grasping

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · 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 a double deep Q-network trained entirely on domain-randomized simulated RGB images can output gripper-centric Cartesian motions that achieve directional (top-down and side) grasps of four specific YCB objects, and…

desk verdict A genuinely new problem framing and a decent sim-to-real demonstration, undercut mainly by a thin real-world evaluation; worth a serious referee. read the letter →

arxiv 1909.02075 v3 pith:VIW76OY4 submitted 2019-09-04 cs.RO

Shariq Iqbal , Jonathan Tremblay , Thang To , Jia Cheng , Erik Leitch , Andy Campbell , Kirby Leung , Duncan McKay
show 1 more author
Stan Birchfield
This is my paper · ORCID
classification cs.RO
keywords directionalsemanticgraspingsim-to-realtransferdomainrandomizationdeepreinforcementlearningdoubleQ-networkcross-entropymethodroboticRGBcamera
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 attempts to show that a robot can learn to grasp a specific object from a specific direction, top-down or side, using only simulated RGB images from a wrist-mounted camera, and that the learned policy transfers to a real robot without any fine-tuning. The authors train a double deep Q-network on domain-randomized, photorealistic simulated data and convert its Q-values into gripper-centric Cartesian motions via the cross-entropy method. If the claim holds, it extends semantic grasping beyond the usual top-down pick of any object to directional grasps that matter for tasks like placing one object next to another. The paper reports promising success rates on four YCB objects in both simulation and the real world, while noting the need for future work on robustness.

What carries the argument

The load-bearing components are: (1) a double deep Q-network (DDQN) that maps a 64×40 RGB image plus a continuous 4D action to a Q-value, with two auxiliary heads predicting gripper-to-object centroid distance and rotational offset; (2) the cross-entropy method (CEM), which iteratively samples candidate actions from a Gaussian, keeps the best Q-valued ones, and refits the distribution to select the final action; (3) a custom Unreal Engine 4 simulator that models physical contact and produces photorealistic, domain-randomized images—random table textures, hue/saturation/brightness jitter, and Gaussian blur—to bridge the reality gap; and (4) a reward that penalizes contact forces on the gripper tips to prevent the policy from exploiting a physics-engine deficiency where the gripper snaps into place around the object.

What would settle it

Train and deploy the same architecture on the same four YCB objects, then run the real-robot evaluation with the Baxter's stock camera settings and unmodified room lighting, including specular surfaces; if success rates for previously successful objects fall near zero, the sim-to-real claim fails outside the manually tuned conditions.

Watch

Extended reading notes

Core claim

The central claim is that a double deep Q-network, trained entirely on domain-randomized simulated monocular RGB images, can output gripper-centric Cartesian motions that achieve directional (top-down and side) grasps of four specific YCB objects (meat can, mustard bottle, sugar box, soup can), and that this policy transfers to the real world without a domain-adaptation step. The action space is 4-DOF: a translation of up to 5 cm and a rotation of up to π/4 rad about the wrist axis, expressed in the gripper coordinate frame so the policy is agnostic to absolute object orientation. In real-world tests the network reached 83.3% top-down and 75% multi-side success for the meat can, 100% top-down and 66.7% multi-side for the mustard bottle, 50% top-down and 66.7% multi-side for the sugar box, and 33.3% top-down and 50% multi-side for the soup can.

Load-bearing premise

The paper's claim of sim-to-real transfer without adaptation rests on the assumption that manually adjusting the real camera gain and room lighting to reduce reflections is a minor preprocessing step rather than a form of environment tuning; if the physical world must be adjusted to match the simulator, the demonstrated transfer is confined to a narrow envelope of conditions.

Editorial extensions

If this is right

  • If the claim is correct, directional semantic grasps can be trained without any real-world grasp data, removing a major data-collection bottleneck.
  • A gripper-centric action space means the learned policy can grasp objects from any absolute orientation, so the same network handles top-down and side grasps without retraining per viewpoint.
  • The simulator's design criteria—contact modelling, photorealistic rendering, and domain randomization—become a template for other sim-to-real reinforcement learning tasks in manipulation.
  • The method still relies on a pre-grasp initialization (from pose estimation, not used in this work), so it is best seen as a grasp controller rather than a full pick-and-place system.

Reading between the lines

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

  • The manual adjustment of camera gain and lighting suggests that domain randomization did not cover specular reflections and sensor noise; a testable extension is to add material-level specular properties and camera noise models to the simulator and re-measure transfer.
  • The success rates vary widely by object and direction (33–100%), implying the method is not yet robust; a neighbouring problem worth exploring is a confidence-based retry or regrasp policy built on the same Q-function.
  • The physics-engine exploit the authors had to penalize away indicates that reinforcement learning can latch onto simulator-specific dynamics; a useful diagnostic is to compare grasp kinematics in simulation versus reality to detect such artifacts before deployment.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper addresses directional semantic grasping: grasping one of four YCB objects from a specified direction (top or side) using monocular RGB images from a wrist-mounted camera. The authors train a double deep Q-network entirely in simulation, with continuous actions (3D translation plus wrist rotation) selected by the cross-entropy method, and with domain randomization applied to textures, color, and blur. Training data are collected off-policy in a custom Unreal Engine-based simulator that models contacts and produces photorealistic images. Evaluation is in a held-out simulated environment and in the real world with a Baxter robot. The central claim is that a policy learned entirely on simulated data transfers to the real world without a domain-adaptation step, with promising but not state-of-the-art success rates, and the authors are explicit about limitations.

Significance. If the transfer claim holds, this is a valuable proof-of-concept: it extends simulation-only RGB training from the single-color objects used in prior work to textured household objects, and it demonstrates directional rather than only top-down grasping. The simulator itself, with contact modeling and domain randomization, is a useful engineering contribution, and the paper's decision to report challenges (physics-engine exploitation, friction mismatches, specular reflections) rather than only positive results is a strength. I find no circularity problem: the auxiliary targets are training signals derived from the same reward, not independent validation, and the paper does not relabel fitted parameters as predictions. The main weakness is that the real-world evidence, as reported, does not yet bear the weight of the central claim because of missing trial counts, a missing open-loop baseline, and manual adjustment of the real camera and lighting.

major comments (3)
  1. [IV-B, Table I] Table I reports real-world grasp success rates as single percentages without trial counts. With small trial counts, 83.3% is consistent with 5/6, 66.7% with 2/3 or 4/6, and 33.3% with 1/3 or 2/6; 100.0% could be as few as one success. The binomial confidence intervals are wide enough that several cells are weak evidence for the reported ordering of object and grasp-direction performance. Provide the number of trials for each cell and report confidence intervals or raw counts, or run enough trials to make the rates informative.
  2. [IV-B and II] Every real-world trial starts from a manual pre-grasp above the object in roughly the same area as in simulation, which is the same narrow initial-condition distribution used in training (Section II: 5x5 cm area, 2-4.5 cm above the object, orientation within ±π/4). The policy then executes only five steps with at most 5 cm of translation each, so the object is already nearly centered and at the correct approach direction at the start. Without an open-loop or image-ablated baseline (e.g., a fixed closing motion, or a policy receiving a blank or occluded image), it is not clear how much of Table I is due to the learned visual closed-loop policy. Add such a baseline to support the attribution of success to the visual policy.
  3. [IV-B] The paper states that the authors had to manually adjust the gain of the camera and adjust room lighting to reduce reflections not modeled in simulation. This means the real-world visual conditions were tuned to match the simulator, which narrows the claimed transfer without domain adaptation to a specific, manually selected envelope. Describe the range and magnitude of these adjustments, and ideally evaluate the policy under unadjusted or varied lighting and gain settings to quantify sensitivity.
minor comments (5)
  1. [IV-A] The held-out simulation evaluation uses lighting and textures that remain consistent throughout the experiments, but the text does not give final per-object success rates with trial counts or confidence intervals; Figure 4 shows training curves without error bars, which makes it hard to judge the variance of the reported '95% success or more' statements.
  2. [III] The simulator is described as supporting both PhysX and FleX, but the paper does not state which physics solver and contact parameters were used for the reported training and evaluation; this information is needed for reproducibility.
  3. [Figure 5] Figure 5 shows representative real-world grasps, but the caption does not specify which object and which grasp direction each row or trial corresponds to; labeling each panel would aid interpretation.
  4. [References] Reference [31] misspells 'Viereck' as 'Vierech', and reference [32] has an extra comma in the author list ('J. Peters, , and G. Neumann').
  5. [II] The displayed equation for the return Rt has garbled formatting for the summation and discount factor; it should be typeset with proper limits and subscripts for readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the real-world grasp rates are an external measurement, not a by-construction consequence of the training inputs.

full rationale

The paper's derivation chain is empirical: a DDQN is trained solely on domain-randomized simulated RGB images, and the central evidence is the measured grasp success in a held-out simulated environment and on a real Baxter robot. No parameter is fitted to the real-world results and then relabeled as a prediction; the real-world table is an external measurement of a policy trained before those trials. The auxiliary targets (distance to object centroid and rotational offset) are training-signal features correlated with the reward, not outputs used as evidence for transfer, so they do not make the central claim circular. The only self-citation that appears in a design decision is the use of the authors' prior pose-estimation work [14] to choose the initial gripper randomization range; this is not load-bearing for the transfer claim because that pose estimator is explicitly not used in the final system and the initialization choice is an external, falsifiable result rather than a component whose value is derived from the success table. The manual gain and lighting adjustments in Section IV-B are a domain-match limitation and a threat to the breadth of the sim-to-real claim, but they are not a circular reduction of the prediction to its inputs. Therefore no circularity is exhibited, and the score is 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim rests on three domain assumptions: simulator fidelity, domain randomization sufficiency, and object representativeness. The free parameters are mostly training hyperparameters and heuristic ranges; none are fitted to the real-world outcome in a circular way. No new physical entities are introduced.

free parameters (4)
  • Reward shaping weights
    Distance penalty, object displacement penalty, and contact force penalty weights are chosen by hand without a sensitivity study; they shape the learned policy and are not reported numerically in the paper.
  • CEM action-selection parameters (na, nb, nn) = training: 16, 5, 2; testing: 64, 6, 3
    These sampling parameters affect the quality of the greedy action approximation and are set heuristically.
  • Action space limits = 5 cm translation, pi/4 rotation
    These limits bound the possible motions and are chosen conservatively based on expected pose error; they constrain the policy.
  • Domain randomization ranges = gripper start offset +/- 5 cm, height 2 to 4.5 cm, rotation +/- pi/4; blur kernel 1-3 px
    These ranges control how much visual variability the network sees and are chosen by hand.
assumptions (4)
  • domain assumption The simulator's contact physics (PhysX/FleX) are a sufficiently faithful model of the real Baxter gripper and rigid objects for the learned policy to transfer.
    Training is entirely in simulation; the authors found and patched one physics exploit, indicating the assumption is only partially satisfied.
  • domain assumption Domain randomization over textures, hue, saturation, brightness, and Gaussian blur is sufficient to bridge the visual reality gap for RGB-only policies.
    The manual adjustment of real lighting and camera gain in Section IV-B suggests this assumption does not fully hold.
  • domain assumption The four selected YCB objects are representative of household objects for which sim-to-real transfer will generalize.
    Stated in Section VI as a conjecture; no evidence is provided for generalization to unseen objects.
  • standard math Standard deep Q-learning convergence and hyperparameter choices are appropriate for this POMDP setting.
    The paper uses DDQN as a standard tool; no formal convergence guarantees are given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Toward Sim-to-Real Directional Semantic Grasping." pith.science (2026). https://pith.science/paper/VIW76OY4

@misc{pith2026190902075,
  author       = {Pith},
  title        = {Pith review of: Toward Sim-to-Real Directional Semantic Grasping},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VIW76OY4}},
  note         = {Machine review of arXiv:1909.02075}
}
read the original abstract

We address the problem of directional semantic grasping, that is, grasping a specific object from a specific direction. We approach the problem using deep reinforcement learning via a double deep Q-network (DDQN) that learns to map downsampled RGB input images from a wrist-mounted camera to Q-values, which are then translated into Cartesian robot control commands via the cross-entropy method (CEM). The network is learned entirely on simulated data generated by a custom robot simulator that models both physical reality (contacts) and perceptual quality (high-quality rendering). The reality gap is bridged using domain randomization. The system is an example of end-to-end (mapping input monocular RGB images to output Cartesian motor commands) grasping of objects from multiple pre-defined object-centric orientations, such as from the side or top. We show promising results in both simulation and the real world, along with some challenges faced and the need for future research in this area.

Figures

Figures reproduced from arXiv: 1909.02075 by the authors.

Figure 1
Figure 1. We have developed a simulator that models physical contact and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Detailed network structure: The RGB image from the gripper [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Domain randomization, including random table texture, is used to [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Grasp success rate, evaluated on a simulated held-out environment, over the course of training for the four different YCB models. Shown are [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: After training in simulation, the policy is able to grasp the intended object in the real world without fine-tuning. Because the action space is [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 36 canonical work pages

  1. [1]

    Dex-Net 1.0: A cloud-based network of 3D objects for robust grasp planning using a multi-armed bandit model with correlated rewards,

    J. Mahler, F. T. Pokorny, B. Hou, M. Roderick, M. Laskey, M. Aubry, K. Kohlhoff, T. Kr ¨oger, J. Kuffner, and K. Goldberg, “Dex-Net 1.0: A cloud-based network of 3D objects for robust grasp planning using a multi-armed bandit model with correlated rewards,” in IEEE International Conference on Robotics and Automation (ICRA) , 2016, pp. 1957–1964

    work page 2016

  2. [2]

    Dex-Net 2.0: Deep learning to plan robust grasps with synthetic point clouds and analytic grasp metrics,

    J. Mahler, J. Liang, S. Niyaz, M. Laskey, R. Doan, X. Liu, J. A. Ojea, and K. Goldberg, “Dex-Net 2.0: Deep learning to plan robust grasps with synthetic point clouds and analytic grasp metrics,” in Robotics: Science and Systems (RSS) , 2017

    2017

  3. [3]

    Learning deep policies for robot bin picking by simulating robust grasping sequences,

    J. Mahler and K. Goldberg, “Learning deep policies for robot bin picking by simulating robust grasping sequences,” in Proceedings of the 1st Annual Conference on Robot Learning (CoRL) , 2017, pp. 515– 524

    work page 2017

  4. [4]

    Dex- Net 3.0: Computing robust robot suction grasp targets in point clouds using a new analytic model and deep learning,

    J. Mahler, M. Matl, X. Liu, A. Li, D. Gealy, and K. Goldberg, “Dex- Net 3.0: Computing robust robot suction grasp targets in point clouds using a new analytic model and deep learning,” in IEEE International Conference on Robotics and Automation (ICRA) , 2018

    work page 2018

  5. [5]

    End- to-end learning of semantic grasping,

    E. Jang, S. Vijaynarasimhan, P. Pastor, J. Ibarz, and S. Levine, “End- to-end learning of semantic grasping,” in CoRL, 2017

    work page 2017

  6. [6]

    Human-level control through deep reinforcement learning,

    V . Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, S. Petersen, C. Beattie, A. Sadik, I. Antonoglou, H. King, D. Kumaran, D. Wierstra, S. Legg, and D. Hassabis, “Human-level control through deep reinforcement learning,” Nature, vol. 518, pp. 529–533, Feb. 2015

    work page 2015

  7. [7]

    Continuous control with deep reinforce- ment learning,

    T. P. Lillicrap, J. J. Hunt, A. Pritzel, N. Heess, T. Erez, Y . Tassa, D. Silver, and D. Wierstra, “Continuous control with deep reinforce- ment learning,” in ICLR, 2016

    work page 2016

  8. [8]

    Deep reinforcement learning with double Q-learning,

    H. van Hasselt, A. Guez, and D. Silver, “Deep reinforcement learning with double Q-learning,” in AAAI, 2016

    work page 2016

Show all 42 references
  1. [9]

    Dropout: A simple way to prevent neural networks from overfitting,

    N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhut- dinov, “Dropout: A simple way to prevent neural networks from overfitting,” The Journal of Machine Learning Research , vol. 15, no. 1, pp. 1929–1958, 2014

    1929

  2. [10]

    Learning hand-eye coordination for robotic grasping with deep learning and large-scale data collection,

    S. Levine, P. Pastor, A. Krizhevsky, J. Ibarz, and D. Quillen, “Learning hand-eye coordination for robotic grasping with deep learning and large-scale data collection,” The International Journal of Robotics Research (IJRR), vol. 37, no. 4–5, pp. 421–436, 2018

    2018

  3. [11]

    Deep reinforcement learning for vision-based robotic grasping: A simulated comparative evaluation of off-policy methods,

    D. Quillen, E. Jang, O. Nachum, C. Finn, J. Ibarz, and S. Levine, “Deep reinforcement learning for vision-based robotic grasping: A simulated comparative evaluation of off-policy methods,” in ICRA, 2018

    2018

  4. [12]

    A tutorial on the cross-entropy method,

    P.-T. De Boer, D. P. Kroese, S. Mannor, and R. Y . Rubinstein, “A tutorial on the cross-entropy method,” Annals of Operations Research , vol. 134, no. 1, pp. 19–67, 2005

    2005

  5. [13]

    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 IROS, 2017

    2017

  6. [14]

    Deep object pose estimation for semantic robotic grasping of household objects,

    J. Tremblay, T. To, B. Sundaralingam, Y . Xiang, D. Fox, and S. Birch- field, “Deep object pose estimation for semantic robotic grasping of household objects,” in CoRL, 2018

    2018

  7. [15]

    MuJoCo: A physics engine for model-based control,

    T. E. Emanuel Todorov and Y . Tassa, “MuJoCo: A physics engine for model-based control,” in IROS, 2012

    2012

  8. [16]

    V-REP: A versatile and scalable robot simulation framework,

    E. Rohmer, S. P. N. Singh, and M. Freese, “V-REP: A versatile and scalable robot simulation framework,” in IROS, 2013

    2013

  9. [17]

    PyBullet: A Python module for physics simulation in robotics, games and machine learning,

    E. Coumans and Y . Bai, “PyBullet: A Python module for physics simulation in robotics, games and machine learning,” 2017. [Online]. Available: http://pybullet.org

    2017

  10. [18]

    Unity: A general platform for intelligent agents,

    A. Juliani, V .-P. Berges, E. Vckay, Y . Gao, H. Henry, M. Mattar, and D. Lange, “Unity: A general platform for intelligent agents,” in arXiv:1809.02627, 2018

    2018 arXiv

  11. [19]

    Simulation tools for model-based robotics: Comparison of Bullet, Havok, MuJoCo, ODE and PhysX,

    T. Erez, Y . Tassa, and E. Todorov, “Simulation tools for model-based robotics: Comparison of Bullet, Havok, MuJoCo, ODE and PhysX,” in ICRA, 2015

    2015

  12. [20]

    The YCB object and model set: Towards common benchmarks for manipulation research,

    B. Calli, A. Walsman, A. Singh, S. Srinivasa, P. Abbeel, and A. M. Dollar, “The YCB object and model set: Towards common benchmarks for manipulation research,” in Intl. Conf. on Advanced Robotics (ICAR), 2015

    2015

  13. [21]

    Benchmarking in manipulation research: Using the Yale-CMU- Berkeley object and model set,

    ——, “Benchmarking in manipulation research: Using the Yale-CMU- Berkeley object and model set,” IEEE Robotics and Automation Magazine, vol. 22, no. 3, Sept. 2015

    2015

  14. [22]

    Closing the loop for robotic grasping: A real-time, generative grasp synthesis approach,

    D. Morrison, P. Corke, and J. Leitner, “Closing the loop for robotic grasping: A real-time, generative grasp synthesis approach,” in RSS, 2018

    2018

  15. [23]

    Cartman: The low-cost Cartesian manipulator that won the Amazon Robotics Challenge,

    D. Morrison, A. W. Tow, M. McTaggart, R. Smith, N. Kelly-Boxall, S. Wade-McCue, J. Erskine, R. Grinover, A. Gurman, T. Hunn, D. Lee, A. Milan, T. Pham, G. Rallos, A. Razjigaev, T. Rowntree, K. Vijay, Z. Zhuang, C. Lehnert, I. Reid, P. Corke, and J. Leitner, “Cartman: The low-c...

    2018

  16. [24]

    Multi- task domain adaptation for deep learning of instance grasping from simulation,

    K. Fang, Y . Bai, S. Hinterstoisser, and M. Kalakrishnan, “Multi- task domain adaptation for deep learning of instance grasping from simulation,” in arXiv:1710.06422, 2017

    2017 arXiv

  17. [25]

    Using simulation and domain adaptation to improve efficiency of deep robotic grasping,

    K. Bousmalis, A. Irpan, P. Wohlhart, Y . Bai, M. Kelcey, M. Kalakr- ishnan, L. Downs, J. Ibarz, P. Pastor, K. Konolige, S. Levine, and V . Vanhoucke, “Using simulation and domain adaptation to improve efficiency of deep robotic grasping,” in ICRA, 2018

    2018

  18. [26]

    Adversar- ial discriminative sim-to-real transfer of visuo-motor policies,

    F. Zhang, J. Leitner, Z. Ge, M. Milford, and P. Corke, “Adversar- ial discriminative sim-to-real transfer of visuo-motor policies,” in arXiv:1709.05746, 2017

    2017 arXiv

  19. [27]

    Sim2Real viewpoint invariant visual servoing by recurrent control,

    F. Sadeghi, A. Toshev, E. Jang, and S. Levine, “Sim2Real viewpoint invariant visual servoing by recurrent control,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 4691–4699

    2018

  20. [28]

    Transferring end-to-end visuomotor control from simulation to real world for a multi-stage task,

    S. James, A. J. Davison, and E. Johns, “Transferring end-to-end visuomotor control from simulation to real world for a multi-stage task,” in CoRL, 2017

    2017

  21. [29]

    Sim-to-real reinforce- ment learning for deformable object manipulation,

    J. Matas, S. James, and A. J. Davison, “Sim-to-real reinforce- ment learning for deformable object manipulation,” arXiv:1806.07851, 2018

    2018 arXiv

  22. [30]

    Sim-to-real transfer of accurate grasping with eye-in-hand observations and continuous control,

    M. Yan, I. Frosio, S. Tyree, and J. Kautz, “Sim-to-real transfer of accurate grasping with eye-in-hand observations and continuous control,” in NIPS Workshop on Acting and Interacting in the Real World: Challenges in Robot Learning , 2018

    2018

  23. [31]

    Learning a visuo- motor controller for real world robotic grasping using simulated depth images,

    U. Viereck, A. ten Pas, K. Saenko, and R. Platt, “Learning a visuo- motor controller for real world robotic grasping using simulated depth images,” in CoRL, 2017

    2017

  24. [32]

    Experiments with hierarchical reinforcement learning of multiple grasping policies,

    T. Osa, J. Peters, , and G. Neumann, “Experiments with hierarchical reinforcement learning of multiple grasping policies,” in ISER, 2016

    2016

  25. [33]

    Learning task-oriented grasping for tool manipulation from simulated self-supervision,

    K. Fang, Y . Zhu, A. Garg, A. Kurenkov, V . Mehta, L. Fei-Fei, and S. Savarese, “Learning task-oriented grasping for tool manipulation from simulated self-supervision,” in RSS, 2018

    2018

  26. [34]

    The MOPED framework: Object recognition and pose estimation for manipulation,

    A. Collet, M. Martinez, and S. S. Srinivasa, “The MOPED framework: Object recognition and pose estimation for manipulation,” Interna- tional Journal of Robotics Research , vol. 30, no. 10, pp. 1284–1306, 2011

    2011

  27. [35]

    SimTrack: A simulation-based framework for scalable real-time object pose detection and tracking,

    K. Pauwels and D. Kragic, “SimTrack: A simulation-based framework for scalable real-time object pose detection and tracking,” in IROS, 2015

    2015

  28. [36]

    Goal-directed robot manipulation through axiomatic scene estimation,

    Z. Sui, L. Xiang, O. C. Jenkins, and K. Desingh, “Goal-directed robot manipulation through axiomatic scene estimation,” International Journal of Robotics Research (IJRR) , vol. 36, no. 1, pp. 86–104, 2017

    2017

  29. [37]

    Semantic robot programming for goal-directed manipulation in cluttered scenes,

    Z. Zeng, Z. Zhou, Z. Sui, and O. C. Jenkins, “Semantic robot programming for goal-directed manipulation in cluttered scenes,” in ICRA, 2018

    2018

  30. [38]

    ShapeNet: An information-rich 3D model repository,

    A. X. Chang, T. Funkhouser, L. Guibas, P. Hanrahan, Q. Huang, Z. Li, S. Savarese, M. Savva, H. S. S. Song, J. Xiao, L. Yi, and F. Yu, “ShapeNet: An information-rich 3D model repository,” in arXiv:1512.03012, 2015

    2015 arXiv

  31. [39]

    Beyond PASCAL: A bench- mark for 3D object detection in the wild,

    Y . Xiang, R. Mottaghi, and S. Savarese, “Beyond PASCAL: A bench- mark for 3D object detection in the wild,” in IEEE Winter Conference on Applications of Computer Vision (WACV) , 2014

    2014

  32. [40]

    Mechanical search: Multi-step retrieval of a target object occluded by clutter,

    M. Danielczuk, A. Kurenkov, A. Balakrishna, M. Matl, D. Wang, R. Martn-Martn, A. Garg, S. Savarese, and K. Goldberg, “Mechanical search: Multi-step retrieval of a target object occluded by clutter,” in ICRA, 2019

    2019

  33. [41]

    Mastering the game of go with deep neural networks and tree search,

    D. Silver, A. Huang, C. J. Maddison, A. Guez, L. Sifre, G. van den Driessche, J. Schrittwieser, I. Antonoglou, V . Panneershelvam, M. Lanctot, S. Dieleman, D. Grewe, J. Nham, N. Kalchbrenner, I. Sutskever, T. Lillicrap, M. Leach, K. Kavukcuoglu, T. Graepel, and D. Hassabis, “M...

    2016

  34. [42]

    3D simulation for robot arm control with deep Q-learning,

    S. James and E. Johns, “3D simulation for robot arm control with deep Q-learning,” in NIPS Workshop on Deep Learning for Action and Interaction , 2016

    2016

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.