Pith. sign in

REVIEW 4 major objections 5 minor 39 references

UniTac: Whole-Robot Touch Sensing Without Tactile Sensors

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 ↗

Pith's one-line read Using only the joint torque and position sensors already on a robot, UniTac learns to localize touches, reaching average errors of 7.2 cm on a Spot quadruped and 8.0 cm on a Franka arm at about 2,000 Hz.

desk verdict Useful no-hardware contact localization on real robots, but the headline errors only cover a discrete set of training points, so the whole-robot claim needs a continuity test. read the letter →

arxiv 2507.07980 v1 pith:OUOXL5UJ submitted 2025-07-10 cs.RO

Wanjia Fu , Hongyu Li , Ivy X. He , Stefanie Tellex , Srinath Sridhar This is my paper
classification cs.RO
keywords tactilesensingcontactlocalizationproprioceptionjointtorquehuman-robotinteractionquadrupedrobotarmdata-drivenrobotics
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

The paper tries to establish that whole-body touch sensing is possible on ordinary robots that lack tactile skins, using only the joint torque and position sensors already built into commercial platforms. If true, contact localization becomes an off-the-shelf capability: a few hours of labeling where human touches land, and a robot can react to touch in real time without new hardware. The key result is a regression model that maps proprioceptive signals to a 3D contact point, with average errors of 7.2 cm on a Spot quadruped and 8.0 cm on a Franka arm at about 2,000 Hz. The authors also show the model transfers to a second Spot instance without retraining and powers touch-based interaction demonstrations.

What carries the argument

The load-bearing object is UniTac-Net, a four-layer MLP with layer sizes 64, 128, 256, and 128, ReLU hidden activations, dropout 0.3, and a three-output regression head. Its input is the normalized proprioceptive vector of joint positions $q$ and torques $\tau$ (19 DoF on Spot, 7 on Franka), and its output is a predicted contact point $p$ in the robot frame; the no-contact state is coded as $(0,0,0)$, and training uses mean squared error with the Adam optimizer at a fixed learning rate of $2.5 \times 10^{-3}$ for 30 epochs. The mechanism works because different contact locations produce distinguishable joint-torque signatures across randomized poses, letting the network learn the mapping from sensor feedback to body coordinates. An exponential moving average filter with smoothing factor 0.1 and sliding window 40 turns the 2,000 Hz raw predictions into temporally stable live estimates.

What would settle it

Take a Spot or Franka, train UniTac on the paper's sampled contact points, then touch a location between those points (or on a Spot leg, which the paper excluded) while the robot is stationary, with ground truth measured by a motion-capture marker or a reference tactile patch. If the average localization error on these continuous, out-of-distribution touches is substantially larger than the reported 7-8 cm, then the 'whole-body' claim holds only for the discrete, hand-picked contacts the model was trained on.

Watch

Extended reading notes

Core claim

UniTac's central claim is that whole-body touch localization can be learned from proprioception alone: a four-layer MLP maps joint positions and torques to 3D contact coordinates in the robot frame, reaching average L2 errors of 7.2 cm on a Spot quadruped and 8.0 cm on a Franka Research 3 arm at about 2,000 Hz on an RTX 3090 GPU. The authors treat this as a regression problem rather than classification, represent 'no contact' by predicting the origin, and stabilize live outputs with an exponential moving average filter. They further report that the model transfers to a second Spot instance without retraining and demonstrate real-time applications: touch regions on Spot trigger animal-inspired behaviors, and touch points on Franka act as virtual buttons that select which colored block the arm picks up.

Load-bearing premise

The evaluation assumes that touches applied by a human at a small set of hand-picked body points, with the robot stationary and joint poses randomized in a limited range, are representative of every contact the robot will experience; the held-out test points come from the same discrete set and collection protocol, so the reported errors do not by themselves show accurate localization of arbitrary continuous or moving touches.

Editorial extensions

If this is right

  • A robot with no tactile skin can localize a touch on its body to about 7.2 cm on Spot and 8.0 cm on Franka using only joint torque and position data.
  • Regression beats classification by roughly 30 percentage points in accuracy and about 6.5 cm in L2 error, so predicting continuous coordinates is the better design under sensor noise.
  • The model localizes touches on a second, unseen Spot instance without retraining, indicating the learned proprioceptive mapping is not overfit to one robot unit.
  • At roughly 2,000 Hz with an EMA filter, predictions are fast enough for live human-robot interaction such as touch-triggered quadruped choreography and pick-and-place instructions through virtual buttons.
  • Data collection is short (2.5 hours on Spot, 12 minutes on Franka), so equipping a new robot with touch sensing costs labeling time rather than hardware integration.

Reading between the lines

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

  • A natural next test the paper does not run is continuous touch: sliding a finger along an arbitrary path between the sampled points and measuring error with an external tracker; the regression output suggests the model may interpolate, but the reported numbers do not establish it.
  • Because training needs only real-world joint data and a few hours of labeling, the same recipe should transfer to other torque-sensing morphologies such as bipeds or mobile manipulators, and to dynamic contacts during motion, both of which are outside the paper's stationary evaluation.
  • The paper's single-contact assumption leaves open the multi-contact case; a detection head or contact-count prior on top of the same proprioceptive features would be the natural extension for collision recovery and safe physical human-robot interaction.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes UniTac, a supervised method that maps joint position and torque readings to 3D contact coordinates using a four-layer MLP, without any tactile hardware. Training data are collected by human touches at 104 sampled points on a Spot quadruped (with the legs excluded) and 10 points on a Franka arm, under randomized joint configurations. On an 80/20 split of samples drawn from these discrete point sets, UniTac reports average L2 localization errors of 7.2 cm on Spot and 8.0 cm on Franka, at roughly 2,000 Hz on an RTX 3090 GPU. The paper also presents qualitative pHRI demonstrations on both platforms and a qualitative cross-instance generalization test on a second Spot robot.

Significance. If the reported localization accuracy held for arbitrary, previously unseen contact locations, UniTac would be a practical and useful contribution to physical human-robot interaction, because it requires no additional sensors and runs at high frequency. The paper's strengths are its real-world data collection on two distinct embodiments, its simple and reproducible network design, and its concrete demonstrations on a quadruped and an arm. However, the current evaluation protocol draws test samples from the same finite set of discrete contact points used as training labels, so the headline errors measure recognition of known locations rather than continuous whole-body localization. The 'whole-robot' claim is also weaker than stated because Spot's legs are excluded from data collection. These gaps are fixable with additional experiments and do not undermine the core supervised-learning formulation itself.

major comments (4)
  1. [§4.2.1, Table 1] The quantitative evaluation only holds out samples from the same discrete set of contact points used as training labels (104 on Spot, 10 on Franka; §4.1.1–4.1.2). The reported L2 errors therefore measure how well the model identifies one of the known training locations, not how well it localizes a contact at an arbitrary, previously unseen body location. Because the regression output is continuous but all ground-truth labels come from the finite sampled set, a model that effectively memorizes the set could still score well on this protocol. I request leave-one-point-out evaluation and/or a test set of touches at locations not in the training set, with continuous ground truth (e.g., motion capture or a tracked probe), and with per-point errors reported. This is the central missing evidence for the 'whole-robot' claim.
  2. [§4.1.1, Figure 3] The paper uses 'whole-robot' and 'whole-body' in the title, abstract, and introduction, but on Spot the legs are explicitly excluded from data collection and evaluation: the Figure 3 caption states that the 104 sampled points 'cover the whole robot except for the legs.' The claim of whole-robot touch sensing therefore overstates the demonstrated coverage. Please either qualify the claims to the covered body regions or extend data collection and evaluation to the legs and any other uncovered surfaces.
  3. [§4.2.1, Eq. (1)] Accuracy is reported at a single threshold epsilon = 12 cm without justification, and all quantitative results are point estimates from one 80/20 split, with no error bars or multiple seeds. Since the central claims rest on these numbers, the paper should report mean and standard deviation over several training seeds and data splits, and should justify the threshold or report threshold-curve statistics (e.g., area under the accuracy-versus-epsilon curve in Figure 6). Without this, the Acc values in Table 1 are not statistically interpretable and the comparison between methods is fragile.
  4. [§4.2.2, Figure 4] The claimed generalization to a different Spot instance is supported only by qualitative results: the text says that sliding-touch localization on an unseen Spot 'exhibit[s] similar accuracies' but no quantitative numbers are given. Given the stated goal of an off-the-shelf interface for other robots, please provide a quantitative evaluation on the second Spot instance using the same protocol and metrics as Table 1, and ideally also with contact locations not used during training. Without such data, the cross-instance generalization claim is not supported.
minor comments (5)
  1. [§3.1, §4.1] The regression model represents the 'no-contact' state by the point (0,0,0), but the data collection and evaluation sections do not clearly describe how no-contact samples are collected and labeled, or whether the reported metrics include them. Clarify the role of no-contact data in training and in the quantitative results.
  2. [Figure 6] Please label the axes and specify whether the accuracy-threshold curves correspond to Spot, Franka, or both. The caption currently says only that accuracy is evaluated at 'various threshold values.'
  3. [§4.1.1] The statement that 'the robot is kept stationary for each sample' is important for understanding the evaluation scope, but it appears only inside the data collection paragraph. State this assumption explicitly in the method overview and in the limitations discussion, since the applications imply the robot may be touched during motion.
  4. [§4.2.2] The sliding-touch qualitative results are promising, but a quantitative trajectory error (for example, comparing the predicted contact path with the actual sliding path over time) would substantially strengthen the claim that the method tracks continuous, moving contacts.
  5. [§5.2.1] The virtual-button demonstration uses stickers at three locations on Franka; please clarify whether these locations coincide with the 10 sampled training points, and if so, whether the demonstration therefore only tests recognition of known locations rather than arbitrary button placement.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: UniTac's core result is a supervised learning evaluation on held-out proprioceptive data, not a derivation that reduces to its inputs.

full rationale

UniTac learns a mapping from joint positions/torques to contact coordinates using an MSE regression loss against manually recorded ground-truth touch locations (Sec. 3.1, 4.1). The headline L2 errors (7.2 cm on Spot, 8.0 cm on Franka) are computed on a 20% validation split of samples collected under the same protocol (Sec. 4.2.1). This is ordinary supervised learning with held-out evaluation, so predictions are not equivalent to the training labels by construction. Although the test samples come from the same discrete set of 104 (Spot) or 10 (Franka) hand-selected points, that is a limitation on continuous generalization, not circularity: the regression target is a continuous 3D coordinate, the model is not defined in terms of the reported error, and no fitted parameter is renamed as a prediction. Self-citations (e.g., Li et al. 2023/2025a/b in the introduction) are illustrative examples of tactile-sensing applications and are not load-bearing for the method. There is no imported uniqueness theorem, no ansatz smuggled via citation, and no renaming of a known result. The paper therefore does not exhibit a circular derivation chain.

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

The paper is an empirical supervised-learning system, so it introduces no new entities and little formal circularity. Its validity rests on physical and protocol assumptions: contacts produce identifiable joint-torque signatures, quasi-static conditions hold, ground-truth labels are correct, and a discrete set of sampled touch points represents the full contact distribution. Several evaluation and training choices (contact point counts, accuracy threshold, EMA parameters, network hyperparameters) are hand-set and not ablated.

free parameters (5)
  • Number of sampled contact points = 104 (Spot), 10 (Franka)
    Defines output resolution and evaluation coverage; hand-chosen in Sec. 4.1.
  • Accuracy threshold epsilon = 12
    Used in Eq. 1 to define accuracy; no justification or ablation for the value.
  • EMA smoothing factor and sliding window length = 0.1, 40
    Temporal filter settings in Sec. 3.4; hand-chosen.
  • MLP architecture, dropout, learning rate, epochs, batch size = 64-128-256-128, 0.3, 2.5e-3, 30, 256
    Training configuration in Sec. 3; fixed without reported tuning or sensitivity analysis.
  • Number of joint configurations per robot = 50 sets (Spot), 25 sets (Franka)
    Data collection effort in Sec. 4.1; affects pose diversity and generalization.
assumptions (5)
  • domain assumption Different contact locations generate distinguishable proprioceptive feedback in joint torque and position signals.
    Core premise stated in Sec. 1 and used throughout; the entire MLP mapping relies on this separability.
  • domain assumption The robot is in quasi-static conditions during data collection and evaluation (Spot kept stationary, Franka end-effector roughly fixed), so dynamic effects can be ignored.
    Data collection in Sec. 4.1.1 and 4.1.2 keeps robots stationary or near-stationary; the method is not validated during fast motion.
  • domain assumption Ground truth contact labels are accurate; each recorded sample corresponds to a touch at the intended sampled point.
    Labels are assigned by the collection procedure (Sec. 4.1); no independent verification of label accuracy is provided.
  • domain assumption Held-out evaluation on the same discrete sampled point distribution is a valid proxy for general contact localization performance.
    Test split is random from the collected dataset (Sec. 4.2.1); continuous or unseen-location generalization is not measured.
  • domain assumption Joint torque and position sensors on each robot provide sufficient signal-to-noise for localization.
    The method uses only these inputs; no analysis of sensor noise or failure modes is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UniTac: Whole-Robot Touch Sensing Without Tactile Sensors." pith.science (2026). https://pith.science/paper/OUOXL5UJ

@misc{pith2026250707980,
  author       = {Pith},
  title        = {Pith review of: UniTac: Whole-Robot Touch Sensing Without Tactile Sensors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OUOXL5UJ}},
  note         = {Machine review of arXiv:2507.07980}
}
read the original abstract

Robots can better interact with humans and unstructured environments through touch sensing. However, most commercial robots are not equipped with tactile skins, making it challenging to achieve even basic touch-sensing functions, such as contact localization. We present UniTac, a data-driven whole-body touch-sensing approach that uses only proprioceptive joint sensors and does not require the installation of additional sensors. Our approach enables a robot equipped solely with joint sensors to localize contacts. Our goal is to democratize touch sensing and provide an off-the-shelf tool for HRI researchers to provide their robots with touch-sensing capabilities. We validate our approach on two platforms: the Franka robot arm and the Spot quadruped. On Franka, we can localize contact to within 8.0 centimeters, and on Spot, we can localize to within 7.2 centimeters at around 2,000 Hz on an RTX 3090 GPU without adding any additional sensors to the robot. Project website: https://ivl.cs.brown.edu/research/unitac.

Figures

Figures reproduced from arXiv: 2507.07980 by the authors.

Figure 1
Figure 1. Interactions achieved using UniTac. UniTac achieves whole-robot touch sensing without using any tactile sensors and empowers applications such as patting the quadruped for canine-inspired responses or touch-based instructions for manipulation. Our method can be applied to robots with different embodiment types, including quadrupeds and arms. 1. Introduction Commercial robots are becoming increasingly capa￾ble. We no… view at source ↗
Figure 2
Figure 2. Design of UniTac-Net. UniTac-Net is a four￾layer MLP with either a regression or a classification out￾put head. It takes proprioceptive feedback (joint torques and positions) as input and predicts the contact (if any). them as the ground truth contact locations. We col￾lect joint data during contact at each point multiple times by varying joint configurations, and construct a dataset  = {𝑑1, 𝑑2, … , 𝑑𝑘} with 𝑘 samp… view at source ↗
Figure 3
Figure 3. Sampled contact points for data collection. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Live contact localization on Spots. Top row: A human applies touch to the robot. Middle row: The system localizes the contact point on the robot’s mesh. Bottom row: Normalized joint torque changes are displayed (different colors indicate distinct joint sensors). The la…
Figure 5
Figure 5. Figure 5: Live contact localization on FR3 robot arm. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Model accuracy with respect to threshold [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Quadruped interactions design. Touch-based interactions on Spot are categorized into three response types: motion actions (equine-inspired), posture changes, and body expressions (both canine-inspired) 4.2.2. Qualitative Results We demonstrate qualitative results of re…
Figure 8
Figure 8. Figure 8: pHRI deployment on Spot. The first row illustrates the inspiration from human-animal interactions, showcasing how dogs and horses respond to touch cues (colored dots). The second row depicts the corresponding robotic responses in deployment on Spot. Video illustrations…
Figure 9
Figure 9. Figure 9: Robot arm interactions design. With UniTac, we can assign “virtual buttons” for operators to provide manipulation instructions at no cost [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: pHRI deployment on Franka. Each row illustrates Franka picking up a block of a different color. • Lying Down is triggered by touching the middle section, causing Spot to fully lower itself. • Sitting occurs when the rear section near the hip is touched, prompting Spot…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

39 extracted references · 39 canonical work pages

  1. [1]

    In Paul McGreevy, editor, Equine Behavior , pages 151--163

    Chapter 6 - Communication . In Paul McGreevy, editor, Equine Behavior , pages 151--163. W.B. Saunders, Oxford, January 2004

    work page 2004

  2. [2]

    Legged Locomotion in Challenging Terrains using Egocentric Vision , November 2022

    Ananye Agarwal, Ashish Kumar, Jitendra Malik, and Deepak Pathak. Legged Locomotion in Challenging Terrains using Egocentric Vision , November 2022

    work page 2022

  3. [3]

    Argall and Aude G

    Brenna D. Argall and Aude G. Billard. A survey of Tactile Human – Robot Interactions . Robotics and Autonomous Systems, 58 0 (10): 0 1159--1176, October 2010

    work page 2010

  4. [4]

    AnySkin : Plug -and-play Skin Sensing for Robotic Touch , September 2024

    Raunaq Bhirangi, Venkatesh Pattabiraman, Enes Erciyes, Yifeng Cao, Tess Hellebrekers, and Lerrel Pinto. AnySkin : Plug -and-play Skin Sensing for Robotic Touch , September 2024

    work page 2024

  5. [5]

    Block, Sammy Christen, Roger Gassert, Otmar Hilliges, and Katherine J

    Alexis E. Block, Sammy Christen, Roger Gassert, Otmar Hilliges, and Katherine J. Kuchenbecker. The Six Hug Commandments : Design and Evaluation of a Human - Sized Hugging Robot with Visual and Haptic Perception . In Proceedings of the 2021 ACM / IEEE International Conference on Human - Robot Interaction , pages 380--388, Boulder CO USA, March 2021. ACM

    work page 2021

  6. [6]

    Dahiya, Giorgio Metta, Maurizio Valle, and Giulio Sandini

    Ravinder S. Dahiya, Giorgio Metta, Maurizio Valle, and Giulio Sandini. Tactile Sensing — From Humans to Humanoids . IEEE Transactions on Robotics, 26 0 (1): 0 1--20, February 2010

    work page 2010

  7. [7]

    The path towards contact-based physical human-robot interaction, July 2024

    Mohammad Farajtabar and Marie Charbonneau. The path towards contact-based physical human-robot interaction, July 2024

    work page 2024

  8. [8]

    Human- Robot Interaction Based on a Sensitive Bumper Skin

    Manel Frigola, Alicia Casals, and Josep Amat. Human- Robot Interaction Based on a Sensitive Bumper Skin . In 2006 IEEE / RSJ International Conference on Intelligent Robots and Systems , pages 283--287, October 2006

    work page 2006

Show all 39 references
  1. [9]

    The science of interpersonal touch: an overview

    Alberto Gallace and Charles Spence. The science of interpersonal touch: an overview. Neuroscience & Biobehavioral Reviews, 34 0 (2): 0 246--259, 2010

    2010

  2. [10]

    The Franka Emika Robot : A Reference Platform for Robotics Research and Education

    Sami Haddadin, Sven Parusel, Lars Johannsmeier, Saskia Golz, Simon Gabl, Florian Walch, Mohamadreza Sabaghian, Christoph Jähne, Lukas Hausperger, and Simon Haddadin. The Franka Emika Robot : A Reference Platform for Robotics Research and Education . IEEE Robotics & Automation ...

    2022

  3. [11]

    Robovie: an interactive humanoid robot

    Hiroshi Ishiguro, Tetsuo Ono, Michita Imai, Takeshi Maeda, Takayuki Kanda, and Ryohei Nakatsu. Robovie: an interactive humanoid robot. Industrial Robot: An International Journal, 28 0 (6): 0 498--504, December 2001

    2001

  4. [12]

    Intrinsic sense of touch for intuitive physical human-robot interaction

    Maged Iskandar, Alin Albu-Schäffer, and Alexander Dietrich. Intrinsic sense of touch for intuitive physical human-robot interaction. Science Robotics, August 2024

    2024

  5. [13]

    Iwata and S

    H. Iwata and S. Sugano. Human-robot-contact-state identification based on tactile recognition. IEEE Transactions on Industrial Electronics, 52 0 (6): 0 1468--1477, December 2005

    2005

  6. [14]

    Iwate, H

    H. Iwate, H. Hoshino, T. Morita, and S. Sugano. Human-humanoid physical interaction realizing force following and task fulfillment. In Proceedings. 2000 IEEE / RSJ International Conference on Intelligent Robots and Systems ( IROS 2000) ( Cat . No . 00CH37113 ) , volume 1, page...

    2000

  7. [15]

    Robovie

    T. Kanda, H. Ishiguro, T. Ono, M. Imai, and R. Nakatsu. Development and evaluation of an interactive humanoid robot " Robovie ". In Proceedings 2002 IEEE International Conference on Robotics and Automation ( Cat . No . 02CH37292 ) , volume 2, pages 1848--1855 vol.2, May 2002

    2002

  8. [16]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A Method for Stochastic Optimization , January 2017

    2017

  9. [17]

    Online touch behavior recognition of hard-cover robot using temporal decision tree classifier

    Seong-Yong Koo, Jong Gwan Lim, and Dong-soo Kwon. Online touch behavior recognition of hard-cover robot using temporal decision tree classifier. September 2008

    2008

  10. [18]

    DIGIT : A Novel Design for a Low - Cost Compact High - Resolution Tactile Sensor With Application to In - Hand Manipulation

    Mike Lambeta, Po-Wei Chou, Stephen Tian, Brian Yang, Benjamin Maloon, Victoria Rose Most, Dave Stroud, Raymond Santos, Ahmad Byagowi, Gregg Kammerer, Dinesh Jayaraman, and Roberto Calandra. DIGIT : A Novel Design for a Low - Cost Compact High - Resolution Tactile Sensor With A...

    2020

  11. [19]

    Digitizing Touch with an Artificial Multimodal Fingertip , November 2024

    Mike Lambeta, Tingfan Wu, Ali Sengul, Victoria Rose Most, Nolan Black, Kevin Sawyer, Romeo Mercado, Haozhi Qi, Alexander Sohn, Byron Taylor, Norb Tydingco, Gregg Kammerer, Dave Stroud, Jake Khatha, Kurt Jenkins, Kyle Most, Neal Stein, Ricardo Chavira, Thomas Craven-Bartle, Eri...

    2024

  12. [20]

    ViHOPE : Visuotactile In - Hand Object 6D Pose Estimation With Shape Completion

    Hongyu Li, Snehal Dikhale, Soshi Iba, and Nawid Jamali. ViHOPE : Visuotactile In - Hand Object 6D Pose Estimation With Shape Completion . IEEE Robotics and Automation Letters, 8 0 (11): 0 6963--6970, November 2023

    2023

  13. [21]

    HyperTaxel : Hyper - Resolution for Taxel - Based Tactile Signals Through Contrastive Learning

    Hongyu Li, Snehal Dikhale, Jinda Cui, Soshi Iba, and Nawid Jamali. HyperTaxel : Hyper - Resolution for Taxel - Based Tactile Signals Through Contrastive Learning . In 2024 IEEE / RSJ International Conference on Intelligent Robots and Systems ( IROS ) , pages 7499--7506, October 2024

    2024

  14. [22]

    ViTa - Zero : Zero -shot Visuotactile Object 6D Pose Estimation , April 2025 a

    Hongyu Li, James Akl, Srinath Sridhar, Tye Brady, and Taskin Padir. ViTa - Zero : Zero -shot Visuotactile Object 6D Pose Estimation , April 2025 a

    2025

  15. [23]

    V- HOP : Visuo - Haptic 6D Object Pose Tracking , February 2025 b

    Hongyu Li, Mingxi Jia, Tuluhan Akbulut, Yu Xiang, George Konidaris, and Srinath Sridhar. V- HOP : Visuo - Haptic 6D Object Pose Tracking , February 2025 b

    2025

  16. [24]

    Contact Localization for Robot Arms in Motion without Torque Sensing

    Jacky Liang and Oliver Kroemer. Contact Localization for Robot Arms in Motion without Torque Sensing . In 2021 IEEE International Conference on Robotics and Automation ( ICRA ) , pages 6322--6328, May 2021

    2021

  17. [25]

    Localizing external contact using proprioceptive sensors: The Contact Particle Filter

    Lucas Manuelli and Russ Tedrake. Localizing external contact using proprioceptive sensors: The Contact Particle Filter . In 2016 IEEE / RSJ International Conference on Intelligent Robots and Systems ( IROS ) , pages 5062--5069, October 2016

    2016

  18. [26]

    Evaluating the human likeness of an android by comparing gaze behaviors elicited by the android and a person

    TAKASHI MINATO, MICHIHIRO SHIMADA, SHOJI ITAKURA, KANG LEE, and HIROSHI ISHIGURO. Evaluating the human likeness of an android by comparing gaze behaviors elicited by the android and a person. Advanced robotics : the international journal of the Robotics Society of Japan, 20 0 ...

    2006

  19. [27]

    Robovie- IV : A Communication Robot Interacting with People Daily in an Office

    Noriaki Mitsunaga, Takahiro Miyashita, Hiroshi Ishiguro, Kiyoshi Kogure, and Norihiro Hagita. Robovie- IV : A Communication Robot Interacting with People Daily in an Office . In 2006 IEEE / RSJ International Conference on Intelligent Robots and Systems , pages 5066--5072, October 2006

    2006

  20. [28]

    Haptic Communication Between Humans and Robots

    Takahiro Miyashita, Taichi Tajika, Hiroshi Ishiguro, Kiyoshi Kogure, and Norihiro Hagita. Haptic Communication Between Humans and Robots . In Sebastian Thrun, Rodney Brooks, and Hugh Durrant-Whyte, editors, Robotics Research , volume 28, pages 525--536. Springer Berlin Heidelb...

    2007

  21. [29]

    Sukhatme

    Artem Molchanov, Oliver Kroemer, Zhe Su, and Gaurav S. Sukhatme. Contact localization on grasped objects using tactile sensing. In 2016 IEEE / RSJ International Conference on Intelligent Robots and Systems ( IROS ) , pages 216--222, October 2016

    2016

  22. [30]

    Learning Quadruped Locomotion Using Differentiable Simulation , October 2024

    Yunlong Song, Sangbae Kim, and Davide Scaramuzza. Learning Quadruped Locomotion Using Differentiable Simulation , October 2024

    2024

  23. [31]

    Reducing influence of robot’s motion on tactile sensor based on partially linear model

    Taichi Tajika, Takahiro Miyashita, Hiroshi Ishiguro, and Norihiro Hagita. Reducing influence of robot’s motion on tactile sensor based on partially linear model. In 2008 IEEE / RSJ International Conference on Intelligent Robots and Systems , pages 512--517, September 2008

    2008

  24. [32]

    Tiwana, Stephen J

    Mohsin I. Tiwana, Stephen J. Redmond, and Nigel H. Lovell. A review of tactile sensing technologies with applications in biomedical engineering. Sensors and Actuators A: Physical, 179: 0 17--31, June 2012

    2012

  25. [33]

    Quadrupedal Locomotion on Uneven Terrain With Sensorized Feet

    Giorgio Valsecchi, Ruben Grandia, and Marco Hutter. Quadrupedal Locomotion on Uneven Terrain With Sensorized Feet . IEEE Robotics and Automation Letters, 5 0 (2): 0 1548--1555, April 2020

    2020

  26. [34]

    Recent Progress in Tactile Sensing and Machine Learning for Texture Perception in Humanoid Robotics

    Longteng Yu and Dabiao Liu. Recent Progress in Tactile Sensing and Machine Learning for Texture Perception in Humanoid Robotics . Interdisciplinary Materials, n/a 0 (n/a), December 2024

    2024

  27. [35]

    Enable natural tactile interaction for robot dog based on large-format distributed flexible pressure sensors

    Lishuang Zhan, Yancheng Cao, Qitai Chen, Haole Guo, Jiasi Gao, Yiyue Luo, Shihui Guo, Guyue Zhou, and Jiangtao Gong. Enable natural tactile interaction for robot dog based on large-format distributed flexible pressure sensors. arXiv preprint arXiv:2303.07595, 2023

    2023 arXiv

  28. [36]

    In-situ Mechanical Calibration for Vision -based Tactile Sensors

    Can Zhao, Jieji Ren, Hexi Yu, and Daolin Ma. In-situ Mechanical Calibration for Vision -based Tactile Sensors . In 2023 IEEE International Conference on Robotics and Automation ( ICRA ) , pages 10387--10393, May 2023

    2023

  29. [37]

    Dynamics- Oriented Underwater Mechanoreception Interface for Simultaneous Flow and Contact Perception

    Hua Zhong, Yaxi Wang, Jiahao Xu, Yu Cheng, Sicong Liu, Jia Pan, Wenping Wang, and Zheng Wang. Dynamics- Oriented Underwater Mechanoreception Interface for Simultaneous Flow and Contact Perception . Advanced Intelligent Systems, 7 0 (1): 0 2400492, 2025

    2025

  30. [38]

    Contact Point Localization for Articulated Manipulators with Proprioceptive Sensors and Machine Learning

    Adrian Zwiener, Christian Geckeler, and Andreas Zell. Contact Point Localization for Articulated Manipulators with Proprioceptive Sensors and Machine Learning . In 2018 IEEE International Conference on Robotics and Automation ( ICRA ) , pages 323--329, May 2018

    2018

  31. [39]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

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