Pith. sign in

REVIEW 4 major objections 5 minor 43 references

MISCGrasp: Leveraging Multiple Integrated Scales and Contrastive Learning for Enhanced Volumetric Grasping

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper claims that multi-scale feature fusion with positive-sample contrastive learning lets one volumetric grasping network produce both power and pinch grasps, beating VGN by 25.5 percentage points in declutter rate on pinch-heavy…

desk verdict A plausible architecture-level improvement over VGN, but the headline 25.5% number conflates dataset gains with method gains; needs a fairer baseline comparison and error bars before the claims hold up. read the letter →

arxiv 2507.02672 v1 pith:C46MMESR submitted 2025-07-03 cs.RO cs.CV

Qingyu Fan , Yinghao Cai , Chao Li , Chunting Jiao , Xudong Zheng , Tao Lu , Bin Liang , Shuo Wang This is my paper · ORCID
classification cs.ROcs.CV
keywords volumetricgrasping6-DoFgraspdetectionmulti-scalefeaturescontrastivelearningpinchpowerTSDFsparsesupervision
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

Robotic grasping research has mostly optimized power grasps, where the gripper wraps around the object, leaving pinch grasps on edges and protrusions poorly served. The paper introduces MISCGrasp, a single volumetric network intended to self-adapt between the two grasp types. The authors argue that high-level convolutional features lose the fine geometry needed for pinch grasps, so MISCGrasp fuses a feature pyramid with two transformers that re-inject fine detail while preserving global structure, and adds a positive-sample contrastive loss that aligns multi-scale features of good grasps across the dataset. On tabletop decluttering tasks, including a physical robot setup, they report that MISCGrasp outperforms VGN, with a 25.5-percentage-point gain in declutter rate over VGN in the pinch-focused EGAD+Adv-Pile simulation and a 95.0% versus 37.0% declutter rate in the real-world pile test. They also report that retraining VGN on the new EGAD-based dataset improves VGN itself, so the contribution is partly data and partly architecture.

What carries the argument

The load-bearing machinery is a three-level feature pyramid whose outputs are reconciled by two transformers before decoding. The Insight Transformer operates bottom-up: the high-level feature volume is used as a query, a channel attention map is computed from the low-level volume, and the low-level volume is downsampled and added back, so fine geometric details that deep convolutions would wash out are reinserted into the high-level representation. The Empower Transformer operates on the highest-level volume alone, splitting queries and keys into parts and mixing their softmax attention contributions with learned coefficients, which the paper says prevents overemphasis on self-attention and feature redundancy. After these two transformers, the refined features are concatenated and decoded into voxel-wise grasp quality, orientation, and width. A separate contrastive module interpolates embeddings at all feature levels at the positions of positive grasp labels, projects their concatenation with an MLP, stores them in a nearest-neighbor memory bank, and applies a cosine-similarity loss, so good grasps that are geometrically similar stay close in feature space at every scale.

What would settle it

Run the five-keypoint TSDF pre-pruning and full physics grasp trials on EGAD objects with thin protrusions and narrow cavities, and compare the positive/negative labels. Then retrain both MISCGrasp and VGN on full-trial labels; if the 25.5-point gap shrinks, or if the pre-pruning disagrees with physics trials specifically on pinch-grasp geometries, the training labels rather than the multi-scale architecture carry the reported gain.

Watch

Extended reading notes

Core claim

The paper's central claim is that the reason existing volumetric grasp networks miss pinch grasps is the loss of fine geometric detail in deeper layers, and that this can be repaired by explicit multi-scale fusion plus contrastive alignment of positive samples. Its MISCGrasp network uses a three-level FPN and two modules: an Insight Transformer that lets high-level features query low-level feature volumes for fine detail, and an Empower Transformer that applies a Mixture-of-Softmax-constrained self-attention to the highest-level volume to keep global structure. A contrastive head interpolates features at each scale at positive grasp positions, concatenates them, projects them into a shared space, and pulls each sample toward its nearest neighbor in a memory bank of positive embeddings. On this basis the authors report a 25.5-point gain in declutter rate over VGN on the pinch-heavy EGAD+Adv-Pile scenario, and a real-world pile declutter rate of 95.0% against VGN's 37.0%. They also show that VGN itself improves when trained on their dataset, which they take as evidence that the dataset's scaled EGAD objects provide the pinch-grasp supervision that earlier datasets lacked.

Load-bearing premise

The load-bearing assumption is that the five-keypoint TSDF collision check used to pre-prune grasp candidates labels positive and negative grasps as accurately as full physics grasp trials would; if it mislabels thin protrusions or narrow cavities, the training data is corrupted and the reported gains over VGN could be artifacts of the labels rather than the multi-scale architecture.

Editorial extensions

If this is right

  • A single sparse-supervised volumetric network can handle both power and pinch grasps without a separate pinch planner, if the reported results reproduce.
  • Training data diversity, specifically object scale and geometric complexity, is a first-order factor in decluttering performance; the paper shows even the baseline VGN improves when trained on the new dataset.
  • The Insight and Empower Transformers do complementary work, with the Insight module mattering most in cluttered scenes and the Empower module mattering most in single-object scenes, so omitting either one degrades overall performance.
  • Positive-only contrastive alignment is preferable to contrastive losses with negatives for this task, since the negative-sample variant performed slightly worse in most experiments in the paper.
  • The simulation gains carry over to physical decluttering, with the method clearing 95.0% of a real-world pile versus 37.0% for VGN under the reported setup.

Reading between the lines

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

  • Beyond the paper's reported experiments, the cleanest separation of data and architecture is to retrain MISCGrasp and VGN on a dataset labeled by full physics grasp trials instead of the five-keypoint TSDF pre-pruning; if the 25.5-point gap shrinks, the pre-pruning approximation, not the multi-scale design, is doing much of the work.
  • The paper excludes point-cloud-based grasp detectors from comparison on input-type grounds, so whether multi-scale contrastive alignment transfers to point-cloud-based or other non-TSDF representations is an open question outside its claims.
  • The random scaling range from 0.65 to 1.7 times gripper width is what injects pinch-grasp examples into training; an ablation that sweeps this range would quantify how much of the gain is data augmentation rather than network design.
  • The positive-only memory-bank objective suggests a testable variant with a small number of hard negatives mined from near-miss grasps, since the paper's naive negative-sample variant added noise; such negatives might strengthen the representation without the observed degradation.
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

4 major / 5 minor

Summary. The paper proposes MISCGrasp, a volumetric 6-DoF grasp detection method built on a multi-scale Feature Pyramid Network with two transformer modules (Insight and Empower) for cross-scale and self-attention, and a contrastive learning module that pulls multi-scale features of positive grasp samples toward their nearest neighbors in a memory bank. The authors generate a new training dataset using EGAD objects with random scaling to cover both power and pinch grasps, and they evaluate in PyBullet simulation on EGAD single-object, pile, and packed scenes as well as in physical experiments with a UR5 and Robotiq gripper. The central claim is that MISCGrasp outperforms VGN, GIGA, and several variant methods, with a reported 25.5% improvement in declutter rate over VGN in pinch-focused experiments.

Significance. The multi-scale feature utilization and positive-sample contrastive learning are plausible and interesting contributions to the sparse-supervision volumetric grasping line of work, and the EGAD-based dataset with random scaling is a useful resource for studying power versus pinch grasps. The paper provides matched comparisons (MISCGrasp versus VGN trained on the same data) that favor MISCGrasp in most metrics, and the physical experiments support real-world viability. However, the headline improvement is overstated because it mixes dataset and architecture effects, the GIGA baseline is admitted to be at a disadvantage, and all simulation results are single-run point estimates without statistical support. The current evidence is suggestive rather than conclusive; with corrected reporting and additional validation, the method could make a solid contribution.

major comments (4)
  1. [Abstract; Section IV-B, Tables I and II] The headline improvement of 25.5% in declutter rate over VGN is not the matched comparison. The number corresponds to Table II (EGAD+ADV-Pile) where MISCGrasp (DR 62.3%) is compared to VGN trained on the regenerated VGN dataset (DR 36.8%), whereas VGN trained on the same MISCGrasp training data reaches DR 48.7%. Thus roughly 11.9 of the 25.5 percentage points come from the dataset change, not from the proposed architecture. The paper itself notes that VGN on the new dataset 'outperforms that trained on the VGN dataset by a large margin,' so the claim conflates dataset and method contributions. Please report the dataset-matched comparison as the central result and clearly decompose the contribution of the dataset and the architecture in the abstract and conclusions.
  2. [Section IV-B, baseline descriptions] The GIGA baseline is not compared on equal terms. The text concedes that GIGA was trained on a smaller dataset than its original design and at TSDF resolution 80 instead of 40. Reporting GIGA's low numbers in Tables I and II as evidence of MISCGrasp's superiority is not valid. Either retrain GIGA under the same data and resolution conditions as the other baselines (with whatever network-depth adjustments are used for VGN), or remove GIGA from the comparative tables and mention it only as an exploratory result with the caveat clearly stated.
  3. [Section IV-B, 'Results' and Tables I and II] All simulation results in Tables I and II are single-run point estimates with no variance, error bars, or significance tests. Several differences are small in absolute terms (e.g., Table II Packed-Packed DR 83.2% vs 82.6% for VGN on the same data), and some ablations go the other way (e.g., the intra-scene variant has a higher Packed-Packed DR of 87.7%). Without repeated seeds or confidence intervals, the claim that MISCGrasp 'outperforms' baselines and variants is not statistically supported. Please provide mean and standard deviation over at least 3-5 seeds, and indicate which differences are significant.
  4. [Section III-B, grasp data generation] The pre-pruning step that replaces full grasp attempts with a five-keypoint TSDF collision check is not validated, and the five keypoints are not specified. The text says 'use TSDF to check for collisions instead of performing full grasp attempts,' which suggests the positive/negative labels are derived from a geometric heuristic rather than physics-based grasp trials. If this approximation is inaccurate for thin protrusions, narrow cavities, or other pinch-grasp geometries, the training labels are corrupted and the reported gains relative to VGN could be artifacts. Please either (a) validate the pre-pruned labels against the original physics-based pipeline (e.g., agreement rate on a held-out set of grasp candidates), or (b) clarify that the pre-pruning only filters obvious collisions and that the final labels are still obtained from full grasping trials, and resolve the contradictory wording in Section III-B.
minor comments (5)
  1. [Section I and Section IV-A] The phrase 'the original implementation of VGN' in the Introduction is imprecise because Section IV-A states that network depths are adjusted and the input resolution is increased for all baselines; please specify exactly which VGN version is being compared.
  2. [Section III-E] The sentence 'We generate 5100 scenes for each scene type' is unclear because the evaluation uses three distinct scene types (Single, Pile, Packed); please specify which training scene types are used and how they map to the evaluation types.
  3. [Table I, caption] The 'Score' column in Table I is not defined in the caption or text; please explain how it is computed (for example, as the sum of the declutter rates).
  4. [Section III-C] The notation for the Insight Transformer uses Q, K, V in a different sense than in the Empower Transformer; please clarify the tensor shapes and the role of Convstride, and define the pooling axes in the AvgPool operation.
  5. [Section IV-C, Table IV] Table IV reports SR as 90.3% (186/206) but DR as 95.0% (190/200); since the denominators differ (grasp attempts versus objects), consider presenting both counts in the same units or adding a footnote to avoid confusion.

Circularity Check

0 steps flagged · score 1.0 of 10

No construction-level circularity: the grasping gains are externally evaluated, with experimental-fairness caveats that do not amount to circularity.

full rationale

No significant circularity found. MISCGrasp is an empirical learning system: the multi-scale FPN, Insight/Empower Transformers, and contrastive loss are trained with sparse grasp labels produced by a physics-based/TSDF pipeline and evaluated on external object sets (EGAD, Berkeley adversarial) against published baselines. The claimed improvements are not derived from the method's definitions; no equation in Section III reduces a prediction to a fitted input. The contrastive memory-bank target is a self-supervised training signal, but the reported result, declutter rate, is measured externally in simulation and on a physical robot, so it is not the training objective renamed as a prediction. There is no load-bearing self-citation: the cited prior work (VGN, GIGA, EGAD, NNCLR, CBAM, MoS) is external, and the paper's ablations compare against variants rather than invoking any self-authored uniqueness theorem. The paper itself admits evaluation caveats: 'The limited performance of GIGA is likely due to the smaller dataset size used in our experiments compared to the one used in GIGA. Additionally, the TSDF grid resolution is set to 80 in our experiments, differing from the 40 used in GIGA.' Also, the headline 25.5% improvement over VGN uses VGN trained on the original VGN dataset, whereas Table II shows VGN trained on the authors' matched dataset reaches 48.7% DR versus MISCGrasp's 62.3%, narrowing but not eliminating the gap. These are experimental-fairness concerns, not construction-level circularity, so they do not raise the circularity score above 1.

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

The paper is an empirical deep-learning method. It relies on standard training assumptions plus specific domain choices: a simplified gripper model for label generation, sparse supervised labels from the VGN pipeline, PyBullet as a sim-to-real proxy, and an NNCLR-style memory bank. The listed hyperparameters are hand-set and not covered by sensitivity analysis.

free parameters (7)
  • contrastive loss weight = 0.5
    Used in L = Lgrasp + 0.5 * Lcontrast; no ablation or sensitivity analysis is reported for this coefficient.
  • TSDF grid resolution = 80
    Increased from 40 in VGN to retain geometric detail; all experiments use this value, and it changes memory and feature granularity.
  • object scaling range = U(0.65,1.7)
    Random scaling relative to gripper opening used in data generation to balance power and pinch grasp samples; no analysis of how this range affects the method.
  • memory bank size = 32768
    Set for the NNCLR-style positive embedding bank; assumed to approximate dataset distribution without ablation.
  • embedding dimension = 128
    Dimension of projection head output and memory bank embeddings; chosen without reported sensitivity analysis.
  • positive/negative BCE weight = 2:1
    Weighted binary cross-entropy with weight 2 for positive labels and 1 for negative; chosen to mitigate imbalance without ablation.
  • ET mixture parts N = 2
    Number of softmax components in Empower Transformer's mixture-of-softmax attention; no ablation reported.
assumptions (5)
  • domain assumption Five-keypoint gripper simplification yields correct grasp labels.
    Sec. III-B pre-prunes grasp poses using TSDF collision checks on five keypoints instead of running full grasp attempts; if keypoint model misclassifies collisions, the sparse ground-truth labels are corrupted.
  • domain assumption The VGN-style sparse grasp trial pipeline provides sufficient supervision.
    The dataset is generated by the pipeline from [4] with modifications (8 positives, 64 negatives per scene); using sparse labels assumes they capture the grasp quality function well enough to train a generalizing model.
  • domain assumption PyBullet physics transfer to the real UR5/Robotiq platform.
    Simulation experiments are the main evidence for design choices; real-world validation uses 20 rounds per condition, so the sim-to-real gap is only partially checked.
  • domain assumption Multi-scale features from an FPN capture grasp-relevant geometry at the needed scales.
    The method's core premise is that low-level features preserve pinch-grasp details and high-level features give context; the paper provides no direct feature analysis, only downstream performance.
  • domain assumption The NNCLR memory bank approximates the full positive-sample distribution.
    Sec. III-D assumes a 32768-entry bank is large enough to approximate the full dataset distribution; nearest-neighbor retrieval from the bank is the contrastive target.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MISCGrasp: Leveraging Multiple Integrated Scales and Contrastive Learning for Enhanced Volumetric Grasping." pith.science (2026). https://pith.science/paper/C46MMESR

@misc{pith2026250702672,
  author       = {Pith},
  title        = {Pith review of: MISCGrasp: Leveraging Multiple Integrated Scales and Contrastive Learning for Enhanced Volumetric Grasping},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C46MMESR}},
  note         = {Machine review of arXiv:2507.02672}
}
read the original abstract

Robotic grasping faces challenges in adapting to objects with varying shapes and sizes. In this paper, we introduce MISCGrasp, a volumetric grasping method that integrates multi-scale feature extraction with contrastive feature enhancement for self-adaptive grasping. We propose a query-based interaction between high-level and low-level features through the Insight Transformer, while the Empower Transformer selectively attends to the highest-level features, which synergistically strikes a balance between focusing on fine geometric details and overall geometric structures. Furthermore, MISCGrasp utilizes multi-scale contrastive learning to exploit similarities among positive grasp samples, ensuring consistency across multi-scale features. Extensive experiments in both simulated and real-world environments demonstrate that MISCGrasp outperforms baseline and variant methods in tabletop decluttering tasks. More details are available at https://miscgrasp.github.io/.

Figures

Figures reproduced from arXiv: 2507.02672 by the authors.

Figure 1
Figure 1. showcases examples of power and pinch grasps. Most existing works [1]–[6] primarily focus on power grasps, 1State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences. 2School of Artificial Intelligence, University of Chinese Academy of Sciences. 3Qiyuan Lab. *This work was supported in part by the National Natural Science Foundation of China under Grants… view at source ↗
Figure 2
Figure 2. Framework of MISCGrasp. MISCGrasp utilizes multi-scale features and contrastive feature enhancement for self-adaptive 6-DoF grasping. The Multi-scale Feature Utilization Module extracts and integrates diverse geometric features from different scales. The Insight Transformer facilitates query￾based interactions, enriching high-level features with fine-grained details from lower-level features, while the Empower Trans… view at source ↗
Figure 4
Figure 4. Comparison of Visualization Results in Simulation Experi￾ments between MISCGrasp and VGN. The visualized grasps are filtered according to a quality score greater than 0.9. The results demonstrate that our method predicts more diverse and feasible grasps compared to VGN. Additionally, our method produces more accurate voxel-wise grasp quality volumes. from multiple observations. Given the differences of the input, di… view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Physical Experiment Objects. (a) Objects used for single-object experiments. (b) Objects used for multi-object experiments. wise fusion with skip connections [42]. It is observed in Tab. II that the variant without multi￾scale contrastive enhancement suffers significan…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 40 canonical work pages

  1. [1]

    Graspnet-1billion: A large- scale benchmark for general object grasping,

    H.-S. Fang, C. Wang, M. Gou, and C. Lu, “Graspnet-1billion: A large- scale benchmark for general object grasping,” in Proceedings of the IEEE/CVF CVPR, 2020, pp. 11 444–11 453

    work page 2020

  2. [2]

    Contact- graspnet: Efficient 6-dof grasp generation in cluttered scenes,

    M. Sundermeyer, A. Mousavian, R. Triebel, and D. Fox, “Contact- graspnet: Efficient 6-dof grasp generation in cluttered scenes,” in 2021 IEEE ICRA. IEEE, 2021, pp. 13 438–13 444

    work page 2021

  3. [3]

    Grasp- ness discovery in clutters for fast and accurate grasp detection,

    C. Wang, H.-S. Fang, M. Gou, H. Fang, J. Gao, and C. Lu, “Grasp- ness discovery in clutters for fast and accurate grasp detection,” in Proceedings of the IEEE/CVF ICCV , 2021, pp. 15 964–15 973

    work page 2021

  4. [4]

    V olumetric grasping network: Real-time 6 dof grasp detection in clutter,

    M. Breyer, J. J. Chung, L. Ott, R. Siegwart, and J. Nieto, “V olumetric grasping network: Real-time 6 dof grasp detection in clutter,” in CoRL. PMLR, 2021, pp. 1602–1611

    work page 2021

  5. [5]

    Synergies between affordance and geometry: 6-dof grasp detection via implicit representations,

    Z. Jiang, Y . Zhu, M. Svetlik, K. Fang, and Y . Zhu, “Synergies between affordance and geometry: 6-dof grasp detection via implicit representations,” RSS, 2021

    work page 2021

  6. [6]

    Orbitgrasp: Se (3)-equivariant grasp learning,

    B. Hu, X. Zhu, D. Wang, Z. Dong, H. Huang, C. Wang, R. Walters, and R. Platt, “Orbitgrasp: Se (3)-equivariant grasp learning,” in CoRL. PMLR, 2024, pp. 2456–2474

    work page 2024

  7. [7]

    Robustness of power grasp,

    X.-Y . Zhang, Y . Nakamura, K. Goda, and K. Yoshimoto, “Robustness of power grasp,” in 1994 IEEE ICRA . IEEE, 1994, pp. 2828–2835

    work page 1994

  8. [8]

    The grasp taxonomy of human grasp types,

    T. Feix, J. Romero, H.-B. Schmiedmayer, A. M. Dollar, and D. Kragic, “The grasp taxonomy of human grasp types,” IEEE Transactions on human-machine systems, vol. 46, no. 1, pp. 66–77, 2015

    work page 2015

Show all 43 references
  1. [9]

    Human grasping database for activities of daily living with depth, color and kinematic data streams,

    A. Saudabayev, Z. Rysbek, R. Khassenova, and H. A. Varol, “Human grasping database for activities of daily living with depth, color and kinematic data streams,” Scientific data, vol. 5, no. 1, pp. 1–13, 2018

    2018

  2. [10]

    Softness-adaptive pinch-grasp strategy using fingertip tactile information of robot hand,

    S. Park and D. Hwang, “Softness-adaptive pinch-grasp strategy using fingertip tactile information of robot hand,” IEEE RAL, vol. 6, no. 4, pp. 6370–6377, 2021

    2021

  3. [11]

    Emergent hand morphology and control from optimizing robust grasps of diverse objects,

    X. Pan, A. Garg, A. Anandkumar, and Y . Zhu, “Emergent hand morphology and control from optimizing robust grasps of diverse objects,” in 2021 IEEE ICRA . IEEE, 2021, pp. 7540–7547

    2021

  4. [12]

    Acronym: A large-scale grasp dataset based on simulation,

    C. Eppner, A. Mousavian, and D. Fox, “Acronym: A large-scale grasp dataset based on simulation,” in 2021 IEEE ICRA . IEEE, 2021, pp. 6222–6227

    2021

  5. [13]

    Egad! an evolved grasping analysis dataset for diversity and reproducibility in robotic manipula- tion,

    D. Morrison, P. Corke, and J. Leitner, “Egad! an evolved grasping analysis dataset for diversity and reproducibility in robotic manipula- tion,” IEEE RAL, vol. 5, no. 3, pp. 4368–4375, 2020

    2020

  6. [14]

    An economic framework for 6-dof grasp detection,

    X.-M. Wu, J.-F. Cai, J.-J. Jiang, D. Zheng, Y .-L. Wei, and W.-S. Zheng, “An economic framework for 6-dof grasp detection,” in ECCV. Springer, 2024, pp. 357–375

    2024

  7. [15]

    Bigbird: A large-scale 3d database of object instances,

    A. Singh, J. Sha, K. S. Narayan, T. Achim, and P. Abbeel, “Bigbird: A large-scale 3d database of object instances,” in 2014 IEEE ICRA . IEEE, 2014, pp. 509–516

    2014

  8. [16]

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

    B. Calli, A. Singh, A. Walsman, S. Srinivasa, P. Abbeel, and A. M. Dollar, “The ycb object and model set: Towards common benchmarks for manipulation research,” in 2015 IEEE International Conference on Advanced Robotics. IEEE, 2015, pp. 510–517

    2015

  9. [17]

    The princeton shape benchmark,

    P. Shilane, P. Min, M. Kazhdan, and T. Funkhouser, “The princeton shape benchmark,” in Proceedings Shape Modeling Applications,

  10. [18]

    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 2016 IEEE ICRA. ...

    2016

  11. [19]

    3dnet: Large- scale object class recognition from cad models,

    W. Wohlkinger, A. Aldoma, R. B. Rusu, and M. Vincze, “3dnet: Large- scale object class recognition from cad models,” in 2012 IEEE ICRA . IEEE, 2012, pp. 5384–5391

    2012

  12. [20]

    The kit object models database: An object model database for object recognition, localization and manipulation in service robotics,

    A. Kasper, Z. Xue, and R. Dillmann, “The kit object models database: An object model database for object recognition, localization and manipulation in service robotics,” IJRR, vol. 31, no. 8, pp. 927–934, 2012

    2012

  13. [21]

    Domain randomization and generative models for robotic grasping,

    J. Tobin, L. Biewald, R. Duan, M. Andrychowicz, A. Handa, V . Kumar, B. McGrew, A. Ray, J. Schneider, P. Welinder et al. , “Domain randomization and generative models for robotic grasping,” in 2018 IEEE/RSJ IROS. IEEE, 2018, pp. 3482–3489

    2018

  14. [22]

    Efficient grasping from rgbd images: Learning using a new rectangle representation,

    Y . Jiang, S. Moseson, and A. Saxena, “Efficient grasping from rgbd images: Learning using a new rectangle representation,” in 2011 IEEE ICRA. IEEE, 2011, pp. 3304–3311

    2011

  15. [23]

    Dimensionality reduction by learning an invariant mapping,

    R. Hadsell, S. Chopra, and Y . LeCun, “Dimensionality reduction by learning an invariant mapping,” in Proceedings of the IEEE/CVF CVPR, vol. 2. IEEE, 2006, pp. 1735–1742

    2006

  16. [24]

    A simple frame- work for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple frame- work for contrastive learning of visual representations,” in ICML. PMLR, 2020, pp. 1597–1607

    2020

  17. [25]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in Proceedings of the IEEE/CVF CVPR, 2020, pp. 9729–9738

    2020

  18. [26]

    Exploring simple siamese representation learn- ing,

    X. Chen and K. He, “Exploring simple siamese representation learn- ing,” in Proceedings of the IEEE/CVF CVPR, 2021, pp. 15 750–15 758

    2021

  19. [27]

    With a little help from my friends: Nearest-neighbor contrastive learning of visual representations,

    D. Dwibedi, Y . Aytar, J. Tompson, P. Sermanet, and A. Zisserman, “With a little help from my friends: Nearest-neighbor contrastive learning of visual representations,” in Proceedings of the IEEE/CVF ICCV, 2021, pp. 9588–9597

    2021

  20. [28]

    6-dof contrastive grasp proposal network,

    X. Zhu, L. Sun, Y . Fan, and M. Tomizuka, “6-dof contrastive grasp proposal network,” in 2021 IEEE ICRA. IEEE, 2021, pp. 6371–6377

    2021

  21. [29]

    A self-supervised contrastive learning method for grasp outcomes prediction,

    C. Liu, B. Huang, Y . Liu, Y . Su, K. Mai, Y . Zhang, Z. Yi, and X. Wu, “A self-supervised contrastive learning method for grasp outcomes prediction,” in 2023 IEEE International Conference on Real-time Computing and Robotics (RCAR) . IEEE, 2023, pp. 737–742

    2023

  22. [30]

    Graspcontrast: Self-supervised contrastive learning with false negative elimination for 6-dof grasp detection,

    W. Wang, H. Zhu, and M. H. Ang, “Graspcontrast: Self-supervised contrastive learning with false negative elimination for 6-dof grasp detection,” in 2024 IEEE/RSJ IROS . IEEE, 2024, pp. 7294–7300

    2024

  23. [31]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in Proceedings of the IEEE CVPR, 2017, pp. 2117–2125

    2017

  24. [32]

    Cbam: Convolutional block attention module,

    S. Woo, J. Park, J.-Y . Lee, and I. S. Kweon, “Cbam: Convolutional block attention module,” in ECCV. Springer, 2018, pp. 3–19

    2018

  25. [33]

    Breaking the softmax bottleneck: A high-rank rnn language model,

    Z. Yang, Z. Dai, R. Salakhutdinov, and W. W. Cohen, “Breaking the softmax bottleneck: A high-rank rnn language model,” in 2018 ICLR, 2018

    2018

  26. [34]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al. , “Pytorch: An imperative style, high-performance deep learning library,” NeurIPS, vol. 32, 2019

    2019

  27. [35]

    Adam: A method for stochastic optimiza- tion,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimiza- tion,” in 2015 ICLR, 2015

    2015

  28. [36]

    Super-convergence: Very fast training of neural networks using large learning rates,

    L. N. Smith and N. Topin, “Super-convergence: Very fast training of neural networks using large learning rates,” in Artificial intelligence and machine learning for multi-domain operations applications , vol. 11006. SPIE, 2019, pp. 369–386

    2019

  29. [37]

    Edge grasp network: A graph-based se (3)-invariant approach to grasp detection,

    H. Huang, D. Wang, X. Zhu, R. Walters, and R. Platt, “Edge grasp network: A graph-based se (3)-invariant approach to grasp detection,” in 2023 IEEE ICRA . IEEE, 2023, pp. 3882–3888

    2023

  30. [38]

    Icgnet: A unified approach for instance-centric grasping,

    R. Zurbr ¨ugg, Y . Liu, F. Engelmann, S. Kumar, M. Hutter, V . Patil, and F. Yu, “Icgnet: A unified approach for instance-centric grasping,” in 2024 IEEE ICRA . IEEE, 2024, pp. 4140–4146

    2024

  31. [39]

    Pybullet, a python module for physics simulation for games, robotics and machine learning,

    E. Coumans and Y . Bai, “Pybullet, a python module for physics simulation for games, robotics and machine learning,” 2016–2021, [Online]. Available: http://pybullet.org

    2016

  32. [40]

    Open3d: A modern library for 3d data processing,

    Q. Zhou, J. Park, and V . Koltun, “Open3d: A modern library for 3d data processing,” CoRR, vol. abs/1801.09847, 2018

    2018 arXiv

  33. [41]

    Learning ambidextrous robot grasping policies,

    J. Mahler, M. Matl, V . Satish, M. Danielczuk, B. DeRose, S. McKinley, and K. Goldberg, “Learning ambidextrous robot grasping policies,” Science Robotics, vol. 4, no. 26, p. eaau4984, 2019

    2019

  34. [42]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in MICCAI. Springer, 2015, pp. 234–241

    2015

  35. [2004]

    IEEE, 2004, pp. 167–178

    2004

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.