Pith. sign in

REVIEW 4 major objections 6 minor 53 references

ThinTact:Thin Vision-Based Tactile Sensor by Lensless Imaging

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A coded mask can replace the lens in a tactile sensor, yielding a 9.6 mm thick device with 2 ms image reconstruction.

desk verdict ThinTact is a genuine new sensor: a 9.6mm lensless tactile sensor that works in real demos, but the quantitative support for the reconstruction pipeline is thinner than the prose. read the letter →

arxiv 2501.09273 v1 pith:5ZHC4RIH submitted 2025-01-16 cs.RO

Jing Xu , Weihang Chen , Hongyu Qian , Dan Wu , Rui Chen This is my paper · ORCID
classification cs.RO
keywords lenslessimagingtactilesensingseparableamplitudemaskreal-timereconstructionDCTfilteroptimizationgeneticalgorithmrobotmanipulation
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

ThinTact replaces the camera lens in a vision-based tactile sensor with a coded amplitude mask, shrinking the sensor to 9.6 mm thick while keeping a sensing field above 200 mm². The paper argues that lensless imaging can deliver tactile images fast enough and sharp enough for robot control: a new non-iterative reconstruction algorithm reconstructs contact scenes in under 2 ms, and a genetically optimized mask recovers image quality that matches iterative optimization methods. If true, this would let robots use thin, large-area touch sensing in confined spaces, and the paper demonstrates the payoff with texture classification at 98.96% accuracy and manipulation tasks on delicate objects.

What carries the argument

The load-bearing object is the separable binary mask $M = (1 \cdot 1^\top + \varphi \varphi^\top)/2$, which splits the lensless measurement into an open term $P_o X Q_o^\top$ and a coding term $P_c X Q_c^\top$. The new reconstruction pipeline uses a learned linear frequency-spatial-domain joint filter $F(Y) = \Phi_s \odot \mathrm{DCT}^{-1}[\Phi_f \odot \mathrm{DCT}(Y)]$ to estimate the open term, subtracts it, and then applies the closed-form solution $\hat{X} = V_P[(\Sigma_P U_P^\top Y_c U_Q \Sigma_Q) \oslash (\sigma_P \sigma_Q^\top + \tau 1 1^\top)] V_Q^\top$. The mask itself is optimized by a genetic algorithm whose fitness is reconstruction quality (SSIM, PSNR, gradient uniformity) on simulated close-up images, so the recovered sensing field stays uniform.

What would settle it

Press a dense grid of indenters with known depths across the entire 16.1 by 12.6 mm sensing field, collect reconstructions, and measure per-region SSIM against ground truth; if errors cluster at the field boundary or on shapes unlike the training scenes, the filter separation assumption, and with it the 2 ms reconstruction, fails.

Watch

Extended reading notes

Core claim

The central claim is that a separable amplitude mask, not a lens, can form the imaging core of a practical tactile sensor. Under the Texas Two-Step model $Y = P_o X Q_o^\top + P_c X Q_c^\top$, the paper shows that a linear filter combining DCT-domain and spatial-domain masks can separate the open (low-frequency) component from the coding component, leaving a term $P_c X Q_c^\top$ that has a closed-form SVD-based solution. Combined with a genetic-algorithm mask optimization that accounts for the short scene-to-CMOS distance, this yields reconstructed tactile images of quality similar to optimization-based reconstruction at roughly 1000-fold lower compute, in less than 2 ms.

Load-bearing premise

The entire reconstruction rests on the learned linear filter reliably separating the mask-free low-frequency part of the measurement from the mask-coded part for every contact shape; the filter is trained only on simulated images from the calibrated system matrices, with no theoretical guarantee it holds for unseen real deformations.

Editorial extensions

If this is right

  • Robot fingertips can now be thinner than human fingers while retaining large-area high-resolution touch, because the lens-thickness constraint is removed.
  • The reconstruction speed (under 2 ms, roughly 1000 times faster than the optimization baseline) makes lensless tactile feedback usable in closed-loop manipulation and force-controlled grasping.
  • The optimized mask extends the usable field of view relative to MLS masks in close-up imaging, so the whole 16.1 by 12.6 mm sensing area yields uniform reconstructions.
  • Depth accuracy improves by about 60 percent when a CycleGAN Real2Sim step cleans the reconstructed images before photometric stereo, reaching a 0.13 mm average error.
  • The sensor can classify 47 fabric textures at 98.96 percent accuracy from single static touches, and it can grasp a 0.5 mm pencil lead while detecting contact at 0.16 N.

Reading between the lines

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

  • Lensless tactile sensing could enable dexterous hands where every phalanx carries a sensing patch, since the mask is a flat chrome-on-glass layer that scales with CMOS area and needs no folded optics.
  • The same frequency-spatial filter idea might transfer to other close-range computational imaging problems, such as thin fingerprint imagers or contactless proximity sensing, where a mask sits very close to the sensor.
  • Because the filter is trained on simulated data generated from calibrated system matrices, a domain-randomized training set with varied noise, illumination, and elastomer stiffness could probe how well the separation generalizes beyond the specific fabricated sensor.
  • The thickness could drop below 5 mm with a thin smartphone-class CMOS, since the paper notes the current CMOS alone is over 4 mm thick.
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 / 6 minor

Summary. This paper introduces ThinTact, a lensless vision-based tactile sensor with a sensing field of 16.1 × 12.6 mm² and a thickness of 9.6 mm. The sensor replaces the conventional lens with a separable binary amplitude mask and a CMOS sensor, imaging a colored elastomer membrane. The authors propose three technical components: a non-iterative reconstruction algorithm that estimates the low-frequency 'open' component via a learned DCT-domain and spatial-domain linear filter and then solves for the scene analytically; a mask optimization procedure based on a genetic algorithm with a simulation-based fitness function; and a calibration algorithm to recover the four system matrices of the T2S model. They validate the system with monitor-scene reconstruction comparisons, lateral resolution tests, depth accuracy measurements with a CycleGAN Real2Sim transfer, fabric texture classification, and robot demonstrations including grasping delicate objects, inserting test tubes, and manipulating daily objects. The central claims are that the sensor is thinner than competing straight-optical-path tactile sensors, reconstructs tactile images in under 2 ms on a GPU, and achieves reconstruction quality similar to an iterative optimization baseline.

Significance. If the claims hold, ThinTact is a meaningful step toward thin, large-area vision-based tactile sensing for confined-space robot manipulation, and the non-iterative reconstruction pipeline is a practical contribution that could enable closed-loop tactile feedback at high rates. The paper reports real hardware experiments, a full calibration and reconstruction pipeline, quantitative depth-error measurements, a 98.96% fabric classification accuracy, and multiple robot demonstrations; these are concrete strengths. The main risk is that the quantitative support for the central reconstruction-quality claim is thin: the learned filter's separation behavior is not tested on held-out scenes, the algorithm comparison uses only two scenes without error bars, and the mask optimization and filter training both rely on the same simulated model. Whether the claimed 'similar quality to optimization-based methods' holds across the sensing field and across contact geometries remains to be established.

major comments (4)
  1. [Sec. III-B1, Eq. (8)-(11)] The learned linear filter F(Y) is the load-bearing component of the entire reconstruction pipeline, but the paper does not provide any generalization analysis for it. The filter parameters Φf and Φs are trained via gradient descent on a virtual dataset generated from the calibrated system matrices, and Sec. III-B1 itself states that Yc contains low-frequency components that intermingle with Yo with a spatially varying mixture ratio. Without a held-out evaluation on unseen contact geometries (e.g., indenters, textures, or marker fields not used in the virtual training set) and a per-region analysis over the sensing field, it is not established that F(Y) reliably separates the two components in real operation. Because every downstream quantity — contact geometry, marker flow, texture classification — is computed from the Yc obtained by this subtraction, this is a central gap. I recommend adding a quantitative study of reconstruction error as a function of contact location and shape, plus an ablation that replaces the learned filter with a simple high-pass or fixed-band filter, to demonstrate that the learned separation is necessary and reliable.
  2. [Sec. V-C, Fig. 11 and Table II] The claim that the proposed algorithm 'has reconstruction qualities similar to the traditional optimization-based method' is supported by exactly two representative scenes with no repeated trials or error bars. The reported SSIM/PSNR numbers (0.511/25.748 vs. 0.504/25.928 for the first scene; 0.368/15.312 vs. 0.373/14.867 for the second) show small differences, but with no distribution over scenes or trials it is impossible to judge whether the quality is statistically 'similar' or systematically worse at some regions. I recommend adding a quantitative comparison over a larger set of scenes (e.g., the monitor scenes used in the mask comparison, multiple textures, or scenes at different positions) and reporting per-scene or per-trial statistics such as mean and standard deviation.
  3. [Sec. III-C, Eq. (12)-(15)] The mask optimization and the reconstruction filter are evaluated on the same model that generated them, which creates a circularity risk. The fitness function fSSIM, fPSNR, fGRAD is computed by simulating the measurement with the direct-generation model of the same system matrices and reconstructing with the proposed closed-form solution (11); the filter Φf and Φs is trained on virtual data generated from the calibrated system matrices of the same mask. The real-hardware comparison against the MLS mask (Sec. V-B) is reassuring, but it contains only three scenes and does not report the spatial distribution of errors that the paper attributes to the mask's larger usable FOV. Please add a boundary-region-specific error analysis or a larger set of real scenes to substantiate the uniformity and FOV advantage.
  4. [Sec. VI-C, Table IV] The depth accuracy numbers are reported as single mean values with no standard deviations, no per-image distributions, and no description of the number of images averaged for each cell. For the two test indenters, the Real2Sim error ranges from 0.065 mm to 0.172 mm across the four indentation depths, which is substantial relative to the smallest indentation depth of 0.25 mm; the claim of 'approximately 60%' improvement is based on only two test indenters. Please report the full error distribution and per-indenter statistics so the reader can assess the reliability of the improvement.
minor comments (6)
  1. [Abstract] The abstract says 'generic algorithm' but the text uses 'genetic algorithm'; please correct the typo.
  2. [Sec. III-B1] The statement that the filter parameters Φf and Φs are difficult to obtain analytically from the system matrices is not justified. Please provide a brief explanation of why a closed-form derivation is not feasible, or cite a reference that discusses this difficulty.
  3. [Sec. III-D, Algorithm 3] Algorithm 3 is described as iterating 'until convergence or max number of iterations reached', but no convergence criterion, the number of iterations used in practice, or the final residual error is reported. Please add these details.
  4. [Sec. V-B, Fig. 9] In the third scene of Fig. 9, the PSNR of the baseline mask (14.867) is higher than that of the proposed mask (14.244), while the SSIM is slightly higher for the proposed mask (0.419 vs 0.373). The text says the proposed mask 'largely improved the reconstruction qualities'; please address this exception explicitly or qualify the claim.
  5. [Sec. VI-B, Fig. 12(d)] The GelSight Mini image shown for comparison in Fig. 12(d) lacks acquisition details (same indenter, indentation depth, and illumination conditions). Please provide these details so the resolution comparison is fair.
  6. [Sec. IV, fabrication step (2)] Please specify whether the chrome film of the mask faces the elastomer or the CMOS sensor; this affects the effective distances z and d used in the imaging model.

Circularity Check

1 steps flagged · score 4.0 of 10

Mask optimization is evaluated on the same images used to fit it, making the reported reconstruction-quality improvement partly in-sample; the rest of the sensor pipeline is self-contained hardware evaluation.

  1. fitted input called prediction [Sec. V-B, Eqs. (12)-(21) and Fig. 9]
    "To evaluate each individual’s fitness, we choose the pure white image Xwhite (the 1st row in Fig. 9) and the reshaped USAF test image XUSAF (the 3rd row in Fig. 9), and the fitness score of gene φ is calculated as [Eq. (21)]. ... In Fig. 9, we present the comparison results. Fig. 9 presents three representative scenes. From qualitative comparison and quantitative metrics, we draw the conclusion that our proposed algorithm has largely improved the reconstruction qualities compared with the MLS-based mask, especially in terms of uniformity."

    The genetic algorithm selects the mask vector φ by maximizing the SSIM and PSNR of the proposed reconstruction on Xwhite and XUSAF via Eq. (21), which includes fSSIM(φ, XUSAF), fSSIM(φ, Xwhite), fPSNR, and fGRAD(φ, Xwhite). The subsequent comparison in Fig. 9 then reports SSIM/PSNR for exactly the pure-white scene (1st row) and the USAF chart (3rd row) and concludes 'largely improved reconstruction qualities.' Because the mask parameters were optimized against these very images and metrics, the improvement on those two rows is an in-sample fit, not an independent prediction. Only the second row (GelSight tactile image) is a scene not used in the fitness function, so the mask-optimization claim is only partially independently verified.

full rationale

The paper's central sensor design and real-hardware demonstrations (texture recognition, manipulation, depth accuracy with held-out indenters) provide independent empirical content, so the work is not wholly circular. However, one concrete reduction exists: the optimized mask is fitted by a genetic algorithm whose fitness function directly scores reconstruction quality on Xwhite and the USAF chart, and the mask comparison in Sec. V-B reports improved SSIM/PSNR on those same two images. That portion of the mask-optimization claim is in-sample by construction. The learned filter F(Y) in Eq. (8) is also trained on a virtual dataset generated from the calibrated system matrices, and the algorithm comparison in Sec. V-C does not state whether the two representative scenes were held out from that training; if the USAF chart and GelSight image were in the training set, the 'similar quality to optimization' claim would also be in-sample. The paper offers no generalization analysis for the filter's separation of open and coding components, but that is a missing guarantee rather than a circular reduction. The load-bearing lensless model and closed-form reconstruction are adopted from external prior work [12], [13], not from a self-citation chain. The Real2Sim approach cites the authors' previous work [14], but the current paper re-validates it on ThinTact with held-out indenters, so this self-citation is not load-bearing. Overall, the circularity is partial and localized; the majority of the sensor's capabilities rest on independent experimentation.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on a learned filter, an optimized mask, estimated geometric distances, and a CycleGAN transfer, all fitted to data or tuned for this specific sensor. The mask vector and filter weights are not disclosed. No new physical entities are introduced.

free parameters (6)
  • Filter matrices Phi_f and Phi_s = Not provided; trained via gradient descent
    Used in Eq. (8) to separate Yo from Yc; their quality determines reconstruction accuracy.
  • Mask vector phi = 770-bit binary sequence, not disclosed
    Defines the separable mask and the system matrices; optimized by genetic algorithm to maximize reconstruction fitness.
  • Regularization coefficient tau = Not specified
    Appears in the closed-form reconstruction Eq. (9)-(11); affects resolution and noise tradeoff.
  • dstripe and sigmoid fading parameters = Not specified; experimentally determined
    Used in Sec. III-C2 to generate system matrices for mask optimization simulation; the fading shape is not derived from optics.
  • Scene-to-mask distance z and mask-to-CMOS distance d = z=4.8 mm, d=1 mm (estimated)
    Determine system matrix slopes in Sec. III-C2 and are inputs to the imaging model.
  • CycleGAN weights for Real2Sim = Trained weights not provided
    Used for depth reconstruction in Sec. VI-C; improves reported depth error by about 60 percent.
assumptions (6)
  • domain assumption Texas Two-Step imaging model Y = Po X Qo^T + Pc X Qc^T holds for the close-up tactile geometry.
    Invoked in Eq. (2) and used throughout; if the separable superposition model breaks down, both the filter and the closed-form reconstruction are invalid.
  • ad hoc to paper The mask is separable and symmetric, with phi1 = phi2 = phi.
    Stated in Sec. III-A after Eq. (1); restricts the mask search space and is not required by the physics, only chosen for computational convenience.
  • domain assumption The open component Yo is predominantly low frequency and can be separated from Yc by a learned linear DCT-spatial filter.
    Sec. III-B1 asserts Yo is low frequency but admits Yc also has low-frequency components; the filter is trained rather than derived, so the separation is an empirical assumption.
  • domain assumption The system matrix striped area fades at boundaries according to a sigmoid, with width dstripe determined experimentally.
    Sec. III-C2 uses this to generate system matrices for GA simulation; the exact fading shape is not derived from optics.
  • standard math The closed-form solution for separable masks from FlatCam remains valid after subtracting the estimated open component.
    Used in Eq. (9)-(11); it is a linear least-squares solution from prior work, valid if the coding component estimate is correct.
  • domain assumption The iterative calibration algorithm converges to the true system matrices.
    Sec. III-D Algorithm 3 repeats until convergence but no convergence proof or criterion is given; results depend on this heuristic.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ThinTact:Thin Vision-Based Tactile Sensor by Lensless Imaging." pith.science (2026). https://pith.science/paper/5ZHC4RIH

@misc{pith2026250109273,
  author       = {Pith},
  title        = {Pith review of: ThinTact:Thin Vision-Based Tactile Sensor by Lensless Imaging},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5ZHC4RIH}},
  note         = {Machine review of arXiv:2501.09273}
}
read the original abstract

Vision-based tactile sensors have drawn increasing interest in the robotics community. However, traditional lens-based designs impose minimum thickness constraints on these sensors, limiting their applicability in space-restricted settings. In this paper, we propose ThinTact, a novel lensless vision-based tactile sensor with a sensing field of over 200 mm2 and a thickness of less than 10 mm.ThinTact utilizes the mask-based lensless imaging technique to map the contact information to CMOS signals. To ensure real-time tactile sensing, we propose a real-time lensless reconstruction algorithm that leverages a frequency-spatial-domain joint filter based on discrete cosine transform (DCT). This algorithm achieves computation significantly faster than existing optimization-based methods. Additionally, to improve the sensing quality, we develop a mask optimization method based on the generic algorithm and the corresponding system matrix calibration algorithm.We evaluate the performance of our proposed lensless reconstruction and tactile sensing through qualitative and quantitative experiments. Furthermore, we demonstrate ThinTact's practical applicability in diverse applications, including texture recognition and contact-rich object manipulation. The paper will appear in the IEEE Transactions on Robotics: https://ieeexplore.ieee.org/document/10842357. Video: https://youtu.be/YrOO9BDMAHo

Figures

Figures reproduced from arXiv: 2501.09273 by the authors.

Figure 1
Figure 1. In this work, we propose ThinTact, a thin vision-based tactile sensor [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Workflow of ThinTact. On the left of this figure is the schematic of ThinTact. It utilizes a separable amplitude mask for imaging. To address the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. In the proposed tactile sensor, the scene is close to the CMOS. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: The pipeline of the proposed non-iterative reconstruction algorithm. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Analysis of system matrices. (a) Randomized system matrix generated [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: CAD model of ThinTact. This figure mainly presents the lensless [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 8
Figure 8. Figure 8: Calibration setup for the lensless imaging subsystem. We use a 5.5” [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 7
Figure 7. Figure 7: Dimensions of ThinTact. (a) ThinTact with all its LEDs lighted. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 9
Figure 9. Figure 9: Comparison of the reconstruction qualities using the proposed optimized mask and the traditional MLS-based mask. The reconstruction algorithm is [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Comparison of (a) the MLS-based mask and (b) the optimized mask [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Comparison of the proposed non-iterative reconstruction algorithm and the traditional optimization-based algorithm [12]. Two representative scenes [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: Resolution test of ThinTact. (a) Optical resolution. (b) Screw for [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 13
Figure 13. Figure 13: Indentation experiment for evaluating the depth reconstruction [PITH_FULL_IMAGE:figures/full_fig_p012_13.png]
Figure 14
Figure 14. Figure 14: RGB photos and reconstructed tactile images of the 47 kinds of fabrics used for fabric texture classification. [PITH_FULL_IMAGE:figures/full_fig_p013_14.png]
Figure 15
Figure 15. Figure 15: Fabric texture classification. (a) Dataset collection setup. (b) [PITH_FULL_IMAGE:figures/full_fig_p013_15.png]
Figure 16
Figure 16. Figure 16: ThinTact grasps two delicate objects. On both sides of each image [PITH_FULL_IMAGE:figures/full_fig_p014_16.png]
Figure 18
Figure 18. Figure 18: ThinTact grasps a flat plate. (a) ThinTact is able to fit into the gap at [PITH_FULL_IMAGE:figures/full_fig_p014_18.png]
Figure 19
Figure 19. Figure 19: Drawer opening experiment. (a) ThinTact successfully opens the [PITH_FULL_IMAGE:figures/full_fig_p015_19.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 44 canonical work pages

  1. [1]

    Tactile dexterity: Manipulation primitives with tactile feedback,

    F. R. Hogan, J. Ballester, S. Dong, and A. Rodriguez, “Tactile dexterity: Manipulation primitives with tactile feedback,” in 2020 IEEE interna- tional conference on robotics and automation (ICRA) . IEEE, 2020, pp. 8863–8869

    work page 2020

  2. [2]

    Cable manipulation with a tactile-reactive gripper,

    Y . She, S. Wang, S. Dong, N. Sunil, A. Rodriguez, and E. Adelson, “Cable manipulation with a tactile-reactive gripper,” The International Journal of Robotics Research, vol. 40, no. 12-14, pp. 1385–1401, 2021

    work page 2021

  3. [3]

    Enhancing generalizable 6d pose tracking of an in-hand object with tactile sensing,

    Y . Liu, X. Xu, W. Chen, H. Yuan, H. Wang, J. Xu, R. Chen, and L. Yi, “Enhancing generalizable 6d pose tracking of an in-hand object with tactile sensing,” IEEE Robotics and Automation Letters , 2023

    work page 2023

  4. [4]

    Transtouch: Learning transparent objects depth sensing through sparse touches,

    L. Bian, P. Shi, W. Chen, J. Xu, L. Yi, and R. Chen, “Transtouch: Learning transparent objects depth sensing through sparse touches,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2023, pp. 9566–9573

    work page 2023

  5. [5]

    Recent progress in tactile sensing and sensors for robotic manipulation: can we turn tactile sensing into vision?

    A. Yamaguchi and C. G. Atkeson, “Recent progress in tactile sensing and sensors for robotic manipulation: can we turn tactile sensing into vision?” Advanced Robotics, vol. 33, no. 14, pp. 661–673, 2019

    work page 2019

  6. [6]

    Hardware technology of vision-based tactile sensor: A review,

    S. Zhang, Z. Chen, Y . Gao, W. Wan, J. Shan, H. Xue, F. Sun, Y . Yang, and B. Fang, “Hardware technology of vision-based tactile sensor: A review,” IEEE Sensors Journal , 2022

    work page 2022

  7. [7]

    Swingbot: Learning physical features from in-hand tactile exploration for dynamic swing-up manipulation,

    C. Wang, S. Wang, B. Romero, F. Veiga, and E. Adelson, “Swingbot: Learning physical features from in-hand tactile exploration for dynamic swing-up manipulation,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2020, pp. 5633–5640

    work page 2020

  8. [8]

    Tactile-rl for insertion: Generalization to objects of unknown geome- try,

    S. Dong, D. K. Jha, D. Romeres, S. Kim, D. Nikovski, and A. Rodriguez, “Tactile-rl for insertion: Generalization to objects of unknown geome- try,” in2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2021, pp. 6437–6443

    2021

Show all 53 references
  1. [9]

    Gelsight: High-resolution robot tactile sensors for estimating geometry and force,

    W. Yuan, S. Dong, and E. H. Adelson, “Gelsight: High-resolution robot tactile sensors for estimating geometry and force,” Sensors, vol. 17, no. 12, p. 2762, 2017

    2017

  2. [10]

    The tactip family: Soft optical tactile sensors with 3d-printed biomimetic morphologies,

    B. Ward-Cherrier, N. Pestell, L. Cramphorn, B. Winstone, M. E. Gian- naccini, J. Rossiter, and N. F. Lepora, “The tactip family: Soft optical tactile sensors with 3d-printed biomimetic morphologies,” Soft robotics, vol. 5, no. 2, pp. 216–227, 2018

    2018

  3. [11]

    Tac3d: A novel vision-based tactile sensor for measuring forces distribution and estimating friction coefficient distribution,

    L. Zhang, Y . Wang, and Y . Jiang, “Tac3d: A novel vision-based tactile sensor for measuring forces distribution and estimating friction coefficient distribution,” arXiv preprint arXiv:2202.06211 , 2022

    2022 arXiv

  4. [12]

    Single-frame 3d fluorescence microscopy with ultraminiature lensless flatscope,

    J. K. Adams, V . Boominathan, B. W. Avants, D. G. Vercosa, F. Ye, R. G. Baraniuk, J. T. Robinson, and A. Veeraraghavan, “Single-frame 3d fluorescence microscopy with ultraminiature lensless flatscope,” Science advances, vol. 3, no. 12, p. e1701548, 2017

    2017

  5. [13]

    Flatcam: Thin, lensless cameras using coded aperture and computation,

    M. S. Asif, A. Ayremlou, A. Sankaranarayanan, A. Veeraraghavan, and R. G. Baraniuk, “Flatcam: Thin, lensless cameras using coded aperture and computation,” IEEE Transactions on Computational Imaging, vol. 3, no. 3, pp. 384–397, 2016

    2016

  6. [14]

    Bidi- rectional sim-to-real transfer for gelsight tactile sensors with cyclegan,

    W. Chen, Y . Xu, Z. Chen, P. Zeng, R. Dang, R. Chen, and J. Xu, “Bidi- rectional sim-to-real transfer for gelsight tactile sensors with cyclegan,” IEEE Robotics and Automation Letters , vol. 7, no. 3, pp. 6187–6194, 2022

    2022

  7. [15]

    Trends and challenges in robot manipulation,

    A. Billard and D. Kragic, “Trends and challenges in robot manipulation,” Science, vol. 364, no. 6446, p. eaat8414, 2019

    2019

  8. [16]

    Tactile sensors for advanced intelligent systems,

    C. Wang, L. Dong, D. Peng, and C. Pan, “Tactile sensors for advanced intelligent systems,” Advanced Intelligent Systems , vol. 1, no. 8, p. 1900090, 2019

    2019

  9. [17]

    Localization and manipulation of small parts using gelsight tactile sensing,

    R. Li, R. Platt, W. Yuan, A. Ten Pas, N. Roscup, M. A. Srinivasan, and E. Adelson, “Localization and manipulation of small parts using gelsight tactile sensing,” in 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems . IEEE, 2014, pp. 3988–3993

    2014

  10. [18]

    Dense tactile force esti- mation using gelslim and inverse fem,

    D. Ma, E. Donlon, S. Dong, and A. Rodriguez, “Dense tactile force esti- mation using gelslim and inverse fem,” in2019 International Conference on Robotics and Automation (ICRA) . IEEE, 2019, pp. 5418–5424

    2019

  11. [19]

    Gelslim 3.0: High-resolution measurement of shape, force and slip in a compact tactile-sensing finger,

    I. H. Taylor, S. Dong, and A. Rodriguez, “Gelslim 3.0: High-resolution measurement of shape, force and slip in a compact tactile-sensing finger,” in 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022, pp. 10 781–10 787

    2022

  12. [20]

    Finger-shaped gelforce: sensor for measuring surface traction fields for robotic hand,

    K. Sato, K. Kamiyama, N. Kawakami, and S. Tachi, “Finger-shaped gelforce: sensor for measuring surface traction fields for robotic hand,” IEEE Transactions on Haptics , vol. 3, no. 1, pp. 37–47, 2009. 16

    2009

  13. [21]

    Design, motivation and evaluation of a full-resolution optical tactile sensor,

    C. Sferrazza and R. D’Andrea, “Design, motivation and evaluation of a full-resolution optical tactile sensor,” Sensors, vol. 19, no. 4, p. 928, 2019

    2019

  14. [22]

    Incipient slip detection method with vision-based tactile sensor based on distribution force and deforma- tion,

    R. Sui, L. Zhang, T. Li, and Y . Jiang, “Incipient slip detection method with vision-based tactile sensor based on distribution force and deforma- tion,” IEEE Sensors Journal , vol. 21, no. 22, pp. 25 973–25 985, 2021

    2021

  15. [23]

    Gelslim: A high-resolution, compact, robust, and calibrated tactile- sensing finger,

    E. Donlon, S. Dong, M. Liu, J. Li, E. Adelson, and A. Rodriguez, “Gelslim: A high-resolution, compact, robust, and calibrated tactile- sensing finger,” in 2018 IEEE/RSJ International Conference on Intel- ligent Robots and Systems (IROS) . IEEE, 2018, pp. 1927–1934

    2018

  16. [24]

    Gelsight wedge: Measur- ing high-resolution 3d contact geometry with a compact robot finger,

    S. Wang, Y . She, B. Romero, and E. Adelson, “Gelsight wedge: Measur- ing high-resolution 3d contact geometry with a compact robot finger,” in 2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2021, pp. 6468–6475

    2021

  17. [25]

    Omnitact: A multi-directional high-resolution touch sensor,

    A. Padmanabha, F. Ebert, S. Tian, R. Calandra, C. Finn, and S. Levine, “Omnitact: A multi-directional high-resolution touch sensor,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 618–624

    2020

  18. [26]

    Soft, round, high resolution tactile fingertip sensors for dexterous robotic manipulation,

    B. Romero, F. Veiga, and E. Adelson, “Soft, round, high resolution tactile fingertip sensors for dexterous robotic manipulation,” in 2020 IEEE International Conference on Robotics and Automation (ICRA) . IEEE, 2020, pp. 4796–4802

    2020

  19. [27]

    Geltip: A finger-shaped optical tactile sensor for robotic manipulation,

    D. F. Gomes, Z. Lin, and S. Luo, “Geltip: A finger-shaped optical tactile sensor for robotic manipulation,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2020, pp. 9903–9909

    2020

  20. [28]

    A soft thumb-sized vision- based sensor with accurate all-round force perception,

    H. Sun, K. J. Kuchenbecker, and G. Martius, “A soft thumb-sized vision- based sensor with accurate all-round force perception,” Nature Machine Intelligence, vol. 4, no. 2, pp. 135–145, 2022

    2022

  21. [29]

    Allsight: A low-cost and high-resolution round tactile sensor with zero-shot learning capability,

    O. Azulay, N. Curtis, R. Sokolovsky, G. Levitski, D. Slomovik, G. Lilling, and A. Sintov, “Allsight: A low-cost and high-resolution round tactile sensor with zero-shot learning capability,” IEEE Robotics and Automation Letters , vol. 9, no. 1, pp. 483–490, 2023

    2023

  22. [30]

    Digit: A novel design for a low-cost compact high-resolution tactile sensor with application to in-hand manipulation,

    M. Lambeta, P.-W. Chou, S. Tian, B. Yang, B. Maloon, V . R. Most, D. Stroud, R. Santos, A. Byagowi, G. Kammerer et al. , “Digit: A novel design for a low-cost compact high-resolution tactile sensor with application to in-hand manipulation,” IEEE Robotics and Automation Letters...

    2020

  23. [31]

    A combined tactile and proximity sensing employing a compound-eye camera,

    K. Shimonomura and H. Nakashima, “A combined tactile and proximity sensing employing a compound-eye camera,” in SENSORS, 2013 IEEE . IEEE, 2013, pp. 1–2

    2013

  24. [32]

    Omsense: An omni tactile sensing principle inspired by compound eyes,

    C. Yu, M. Chen, M. D. I. Reyzabal, J. Back, D. Cao, and H. Liu, “Omsense: An omni tactile sensing principle inspired by compound eyes,” IEEE/ASME Transactions on Mechatronics , 2023

    2023

  25. [33]

    Bionic compound eye-inspired high spatial and sensitive tactile sensor,

    L. Song, H. Zhu, Y . Zheng, M. Zhao, C. A. T. Tee, and F. Fang, “Bionic compound eye-inspired high spatial and sensitive tactile sensor,” IEEE Transactions on Instrumentation and Measurement , vol. 70, pp. 1–8, 2021

    2021

  26. [34]

    Multidimensional tactile sensor with a thin compound eye-inspired imaging system,

    Y . Zhang, X. Chen, M. Y . Wang, and H. Yu, “Multidimensional tactile sensor with a thin compound eye-inspired imaging system,” Soft Robotics, vol. 9, no. 5, pp. 861–870, 2022

    2022

  27. [35]

    A thin format vision- based tactile sensor with a microlens array (mla),

    X. Chen, G. Zhang, M. Y . Wang, and H. Yu, “A thin format vision- based tactile sensor with a microlens array (mla),”IEEE Sensors Journal, vol. 22, no. 22, pp. 22 069–22 076, 2022

    2022

  28. [36]

    Recent advances in lensless imaging,

    V . Boominathan, J. T. Robinson, L. Waller, and A. Veeraraghavan, “Recent advances in lensless imaging,” Optica, vol. 9, no. 1, pp. 1–16, 2022

    2022

  29. [37]

    Diffusercam: lensless single-exposure 3d imaging,

    N. Antipa, G. Kuo, R. Heckel, B. Mildenhall, E. Bostan, R. Ng, and L. Waller, “Diffusercam: lensless single-exposure 3d imaging,” Optica, vol. 5, no. 1, pp. 1–9, 2018

    2018

  30. [38]

    Phlatcam: Designed phase-mask based thin lensless camera,

    V . Boominathan, J. K. Adams, J. T. Robinson, and A. Veeraraghavan, “Phlatcam: Designed phase-mask based thin lensless camera,” IEEE transactions on pattern analysis and machine intelligence, vol. 42, no. 7, pp. 1618–1629, 2020

    2020

  31. [39]

    On- chip fluorescence microscopy with a random microlens diffuser,

    G. Kuo, F. L. Liu, I. Grossrubatscher, R. Ng, and L. Waller, “On- chip fluorescence microscopy with a random microlens diffuser,” Optics express, vol. 28, no. 6, pp. 8384–8399, 2020

    2020

  32. [40]

    Learned reconstructions for practical mask-based lensless imaging,

    K. Monakhova, J. Yurtsever, G. Kuo, N. Antipa, K. Yanny, and L. Waller, “Learned reconstructions for practical mask-based lensless imaging,” Optics express, vol. 27, no. 20, pp. 28 075–28 090, 2019

    2019

  33. [41]

    Flatnet: Towards photorealistic scene reconstruction from lensless measurements,

    S. S. Khan, V . Sundar, V . Boominathan, A. Veeraraghavan, and K. Mi- tra, “Flatnet: Towards photorealistic scene reconstruction from lensless measurements,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 4, pp. 1934–1948, 2020

    1934

  34. [42]

    Robust reconstruction with deep learning to handle model mismatch in lensless imaging,

    T. Zeng and E. Y . Lam, “Robust reconstruction with deep learning to handle model mismatch in lensless imaging,” IEEE Transactions on Computational Imaging, vol. 7, pp. 1080–1092, 2021

    2021

  35. [43]

    Flatnet3d: Intensity and absolute depth from single-shot lensless capture,

    D. Bagadthey, S. Prabhu, S. S. Khan, D. T. Fredrick, V . Boominathan, A. Veeraraghavan, and K. Mitra, “Flatnet3d: Intensity and absolute depth from single-shot lensless capture,” JOSA A, vol. 39, no. 10, pp. 1903– 1912, 2022

    1903

  36. [44]

    Efficiency of coordinate descent methods on huge-scale optimization problems,

    Y . Nesterov, “Efficiency of coordinate descent methods on huge-scale optimization problems,” SIAM Journal on Optimization , vol. 22, no. 2, pp. 341–362, 2012

    2012

  37. [45]

    The discrete cosine transform (dct): theory and appli- cation,

    S. A. Khayam, “The discrete cosine transform (dct): theory and appli- cation,” Michigan State University , vol. 114, no. 1, p. 31, 2003

    2003

  38. [46]

    Mitchell, An introduction to genetic algorithms

    M. Mitchell, An introduction to genetic algorithms . MIT press, 1998

    1998

  39. [47]

    Retrographic sensing for the measurement of surface texture and shape,

    M. K. Johnson and E. H. Adelson, “Retrographic sensing for the measurement of surface texture and shape,” in 2009 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 2009, pp. 1070–1077

    2009

  40. [48]

    End-to-end pixelwise surface normal estimation with convolutional neural networks and shape reconstruction using gelsight sensor,

    J. Li, S. Dong, and E. H. Adelson, “End-to-end pixelwise surface normal estimation with convolutional neural networks and shape reconstruction using gelsight sensor,” in 2018 IEEE International Conference on Robotics and Biomimetics (ROBIO) . IEEE, 2018, pp. 1292–1297

    2018

  41. [49]

    Unpaired image-to-image translation using cycle-consistent adversarial networks,

    J.-Y . Zhu, T. Park, P. Isola, and A. A. Efros, “Unpaired image-to-image translation using cycle-consistent adversarial networks,” in Proceedings of the IEEE international conference on computer vision , 2017, pp. 2223–2232

    2017

  42. [50]

    Gelsight mini datasheet,

    GelSight Inc., “Gelsight mini datasheet,” https://www.gelsight.com/ wp-content/uploads/productsheet/Mini/GS Mini Product Sheet 10.07. 24.pdf

  43. [51]

    Dtact: A vision-based tactile sensor that measures high-resolution 3d geometry directly from darkness,

    C. Lin, Z. Lin, S. Wang, and H. Xu, “Dtact: A vision-based tactile sensor that measures high-resolution 3d geometry directly from darkness,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 10 359–10 366

    2023

  44. [52]

    Generation of gelsight tactile images for sim2real learning,

    D. F. Gomes, P. Paoletti, and S. Luo, “Generation of gelsight tactile images for sim2real learning,” IEEE Robotics and Automation Letters , vol. 6, no. 2, pp. 4177–4184, 2021

    2021

  45. [53]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778. Jing Xu received his Ph.D. in mechanical engi- neering from Tsinghua University, Beijing, ...

    2016

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.