Proprioceptive-visual correspondence enables self-other distinction in humanoid robots

Hongkai Xiong; Shikun Ban; Tianyuan Gao; Wenjun Zeng; Wentao Zhu; Yizhong Ge; Yizhou Wang; Yurun Chen

arxiv: 2606.13222 · v1 · pith:PLVPFIDMnew · submitted 2026-06-11 · 💻 cs.RO · cs.AI

Proprioceptive-visual correspondence enables self-other distinction in humanoid robots

Yurun Chen , Tianyuan Gao , Yizhong Ge , Shikun Ban , Yizhou Wang , Hongkai Xiong , Wenjun Zeng , Wentao Zhu This is my paper

Pith reviewed 2026-06-27 06:51 UTC · model grok-4.3

classification 💻 cs.RO cs.AI

keywords self-other distinctionproprioceptive-visual correspondencehumanoid robotsself-modeling3D body occupancymulti-agent scenesmotion planningmotion retargeting

0 comments

The pith

A humanoid robot learns to distinguish itself from others and build a 3D body model using only proprioceptive-visual correspondence without labels or kinematic models.

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

The paper demonstrates that a humanoid robot can acquire self-other distinction solely by learning consistent matches between its internal joint states and external visual observations, without receiving any identity labels or pre-existing knowledge of its own shape. This learned distinction then allows the robot to construct a predictive model that maps its joint configurations to three-dimensional occupancy of its own body as it moves. A sympathetic reader would care because this process operates in real multi-agent environments containing humans or identical robots, and it directly enables practical capabilities such as reaching targets, planning collision-free motions, and retargeting human movements to the robot. The approach therefore supplies a route to bodily self-representation that supports coordination with other agents in shared physical spaces.

Core claim

Proprioceptive-visual correspondence enables self-other distinction in humanoid robots without any identity labels or kinematic models. Once established, this distinction bootstraps a predictive self-model that maps joint configurations to three-dimensional body occupancy, capturing how the robot's body changes with action. In multi-agent scenes involving humans or morphologically identical robots, the system reliably identifies itself, learns the 3D self-model, and supports downstream tasks including target reaching, collision-aware motion planning, and human-to-robot motion retargeting.

What carries the argument

Proprioceptive-visual correspondence, the learned matching between proprioceptive joint states and visual observations, which separates self from others and then drives construction of the 3D occupancy model.

If this is right

The robot identifies itself reliably in scenes containing humans or morphologically identical robots.
It acquires a predictive 3D self-model linking joint configurations to body occupancy.
The self-model supports target reaching, collision-aware motion planning, and human-to-robot motion retargeting.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same correspondence principle could be tested with additional sensory channels such as touch or audio to strengthen self-model accuracy.
Robots using this method might coordinate more effectively in teams by each maintaining an independent self-model.
The approach could reduce reliance on manual calibration procedures when deploying robots in new environments.

Load-bearing premise

Reliable proprioceptive-visual correspondence can be learned and used to separate self from morphologically identical agents without additional supervision, kinematic priors, or identity signals.

What would settle it

A demonstration in which the robot, after training the correspondence, still cannot reliably separate its own body from an identical robot or human in the same workspace, or in which the resulting 3D occupancy predictions do not match actual body positions during movement.

Figures

Figures reproduced from arXiv: 2606.13222 by Hongkai Xiong, Shikun Ban, Tianyuan Gao, Wenjun Zeng, Wentao Zhu, Yizhong Ge, Yizhou Wang, Yurun Chen.

**Figure 1.** Figure 1: Framework overview. a, A humanoid robot in a shared scene faces two coupled problems: identifying which visible body is its own, and predicting how that body should appear under a given proprioceptive state. b, From only proprioceptive states and visual observations, the proposed framework first distinguishes self from other, then learns a kinematics-free self-model. c, Together, self-other distinction and… view at source ↗

**Figure 2.** Figure 2: Self-other distinction from proprioception. a, Held-out accuracy across methods in two evaluation settings: a human-robot scene and a two-humanoid scene with a morphologically matched distractor. Comparisons include direct fusion training and visionlanguage baselines. b, Qualitative examples in which the robot selects its own candidate mask from a scene containing a distractor; the selected mask is marke… view at source ↗

**Figure 3.** Figure 3: Kinematics-free self-modeling from proprioception. a, Simulated examples in which the learned self-model predicts robot body occupancy under different poses and viewpoints. b, Real-world examples in which the predicted occupancy is projected onto the robot while a human distractor remains excluded. c, Qualitative effect of the percentage of correct pseudo-GT masks on the learned self-model. At high pseudo-… view at source ↗

**Figure 4.** Figure 4: Target reaching and collision-aware motion planning. a, Target reaching. A human places a tabletop night-light at different positions, and the robot reaches and touches it to light it up, shown from the front and top views. b, Collision-aware motion planning through a board with a circular aperture. Starting with the left arm hanging down, the robot raises the arm, passes it through the opening, and reache… view at source ↗

**Figure 5.** Figure 5: Human-to-robot motion retargeting. A human demonstration (upper row) is retargeted to a robot motion sequence (lower row) by optimizing the robot’s joint configuration through the learned self-model. We further generalize from controlling a single arm to whole-body motion transfer through human-to-robot retargeting ( [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗

**Figure 6.** Figure 6: Method overview. a, Self-other distinction compares proprioceptive and visual embeddings, selects the self-mask, and trains the alignment with attention-guided contrastive learning. b, Kinematics-free self-modeling uses the selected self-mask to learn a pose-conditioned density and visibility field through bounded volumetric mask rendering. 4.1 Problem formulation and method overview We study kinematics-f… view at source ↗

read the original abstract

Distinguishing self from others is a prerequisite for social intelligence, yet humanoid robots that increasingly share workspaces with humans still lack this ability. Here we show that a humanoid robot can learn self-other distinction from proprioceptive-visual correspondence, without any identity labels or kinematic models. Once established, this distinction bootstraps a predictive self-model that maps joint configurations to three-dimensional body occupancy, capturing how the robot's body changes with action. In multi-agent scenes involving humans or morphologically identical robots, the system reliably identifies itself, learns a 3D self-model, and supports downstream tasks including target reaching, collision-aware motion planning, and human-to-robot motion retargeting. Together, these results outline a route toward bodily self-representation in robots that act and coordinate alongside others in shared physical environments. Project page: https://euron-zc.github.io/humanoid-self-model/.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows a correspondence-based pipeline for label-free self-other distinction and 3D self-modeling on humanoids, but the distinction step looks fragile when agents share morphology.

read the letter

The main takeaway is that this work uses proprioceptive-visual correspondence to let a humanoid separate itself from others without labels or kinematic models, then builds a predictive 3D occupancy model from joint configurations. They demonstrate this on hardware for reaching, collision-aware planning, and human-to-robot retargeting in multi-agent scenes.

The application to bodily self-representation on physical humanoids is the concrete step forward. Prior correspondence methods exist, but tying them directly to a usable 3D self-model and showing downstream task gains on real robots gives the work some practical grounding. The hardware results are the part that stands out.

The soft spot is the self-other partition itself. When morphologies match exactly, visual and proprioceptive streams become highly similar, and the abstract offers no clear mechanism for resolving the resulting ambiguity without hidden priors or post-processing. If the experiments only test clearly distinguishable agents or rely on timing/viewpoint differences that are not stated, the central claim weakens. The stress-test concern holds up on the given description.

The paper targets robotics groups working on embodied self-models and multi-agent coordination. Readers in that niche would extract value from the task demonstrations even if the distinction method needs tighter validation. It deserves peer review because the hardware pipeline is testable and the idea addresses a real gap, though referees should press on the multi-agent failure modes and any unstated disambiguation steps.

Referee Report

1 major / 2 minor

Summary. The paper claims that a humanoid robot can learn self-other distinction solely from proprioceptive-visual correspondence, without identity labels or kinematic models. This distinction then bootstraps a predictive self-model mapping joint configurations to 3D body occupancy. The system is shown to operate in multi-agent scenes with humans or morphologically identical robots, supporting downstream tasks including target reaching, collision-aware motion planning, and human-to-robot motion retargeting.

Significance. If the central empirical results hold, the work offers a label-free, correspondence-driven route to bodily self-representation that could scale to real shared workspaces. The avoidance of explicit supervision and kinematic priors, combined with demonstrations on identical morphologies, would mark a meaningful step beyond model-based or supervised self-modeling approaches in robotics.

major comments (1)

[Section 4] Section 4 (multi-agent experiments): the claim that proprioceptive-visual correspondence alone produces a reliable self/other partition when morphologies are identical is load-bearing for the central contribution. The manuscript does not report quantitative distinction accuracy, ablation on timing/viewpoint synchronization, or failure cases under near-identical streams, leaving open whether residual ambiguity is resolved without undisclosed priors.

minor comments (2)

[Figure 3] Figure 3 and associated text: the 3D occupancy visualizations would benefit from explicit comparison to ground-truth body meshes or failure-mode examples to clarify the predictive self-model's accuracy.
The project page link is provided but the manuscript does not reference specific supplementary videos or code for the correspondence learning pipeline; adding these would aid reproducibility.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on the multi-agent experiments. We address the single major comment below.

read point-by-point responses

Referee: [Section 4] Section 4 (multi-agent experiments): the claim that proprioceptive-visual correspondence alone produces a reliable self/other partition when morphologies are identical is load-bearing for the central contribution. The manuscript does not report quantitative distinction accuracy, ablation on timing/viewpoint synchronization, or failure cases under near-identical streams, leaving open whether residual ambiguity is resolved without undisclosed priors.

Authors: We agree that the current presentation of Section 4 relies primarily on qualitative success in downstream tasks (self-model learning, reaching, planning, retargeting) rather than direct quantitative metrics on the partition itself. No undisclosed priors are used: the method operates exclusively on raw proprioceptive-visual streams and learns correspondence without morphology, identity, or kinematic information. To strengthen the claim, the revised manuscript will add (i) quantitative self/other distinction accuracy (e.g., frame-wise precision/recall against ground-truth ownership), (ii) ablations that systematically vary temporal and viewpoint synchronization, and (iii) explicit discussion of observed failure modes under near-identical streams. These additions will be placed in Section 4 and the supplementary material. revision: yes

Circularity Check

0 steps flagged

No circularity: claims rest on external proprioceptive-visual streams and learned correspondence without self-referential definitions or fitted predictions.

full rationale

The abstract and described pipeline present a learning process that takes proprioceptive and visual inputs as independent data streams to establish self-other distinction via correspondence, then uses that distinction to train a separate predictive occupancy model. No equations, fitting procedures, or self-citations are referenced in the provided text that would reduce the claimed distinction or self-model to a tautology or renamed input. The derivation chain is therefore self-contained against external sensor data and does not invoke any of the enumerated circularity patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review based solely on abstract; ledger entries are therefore provisional and limited to assumptions stated or implied in the summary.

axioms (1)

domain assumption Proprioceptive-visual correspondence is learnable without identity labels or kinematic models
Central premise of the method as stated in the abstract.

pith-pipeline@v0.9.1-grok · 5694 in / 1115 out tokens · 13435 ms · 2026-06-27T06:51:27.901618+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

18 extracted references · 12 canonical work pages

[1]

Retargeting matters: General motion retargeting for humanoid motion tracking.arXiv preprint arXiv:2510.02252,

Joao Pedro Araujo, Yanjie Ze, Pei Xu, Jiajun Wu, and C Karen Liu. Retargeting matters: General motion retargeting for humanoid motion tracking.arXiv preprint arXiv:2510.02252,

arXiv
[2]

Josh Bongard, Victor Zykov, and Hod Lipson

doi: 10.1037/0012-1649.21.6.963. Josh Bongard, Victor Zykov, and Hod Lipson. Resilient machines through continuous self-modeling.Science, 314(5802):1118–1121,

work page doi:10.1037/0012-1649.21.6.963
[3]

URL https://www.science.org/doi/abs/10.1126/science.1133687

doi: 10.1126/science.1133687. URL https://www.science.org/doi/abs/10.1126/science.1133687. Celia A. Brownell, Stephanie Zerwas, and Geetha B. Ramani. “So big”: The development of body self-awareness in toddlers.Child Development, 78(5):1426–1440,

work page doi:10.1126/science.1133687
[4]

Tangney, Roy F

doi: 10.1111/j. 1467-8624.2007.01075.x. Nicolas Carion, Laura Gustafson, Yuan-Ting Hu, Shoubhik Debnath, Ronghang Hu, Didac Suris, Chaitanya Ryali, Kalyan Vasudev Alwala, Haitham Khedr, Andrew Huang, et al. Sam 3: Segment anything with concepts.arXiv preprint arXiv:2511.16719,

work page doi:10.1111/j 2007
[5]

Shaun Gallagher

doi: 10.1016/j.tics.2003.10.004. Shaun Gallagher. Philosophical conceptions of the self: implications for cognitive science. Trends in cognitive sciences, 4(1):14–21,

work page doi:10.1016/j.tics.2003.10.004 2003
[6]

Nerf: Neural radiance field in 3d vision, a comprehensive review.arXiv preprint arXiv:2210.00379,

Kyle Gao, Yina Gao, Hongjie He, Dening Lu, Linlin Xu, and Jonathan Li. Nerf: Neural radiance field in 3d vision, a comprehensive review.arXiv preprint arXiv:2210.00379,

Pith/arXiv arXiv
[7]

17 Yuhang Hu, Boyuan Chen, and Hod Lipson

doi: 10.1109/TMECH.2025.3579247. 17 Yuhang Hu, Boyuan Chen, and Hod Lipson. Egocentric visual self-modeling for autonomous robot dynamics prediction and adaptation.npj Robotics, 3, 06 2025a. doi: 10.1038/ s44182-025-00031-6. Yuhang Hu, Jiong Lin, and Hod Lipson. Teaching robots to build simulations of themselves. Nature Machine Intelligence, 7(3):484–494,...

work page doi:10.1109/tmech.2025.3579247 2025
[8]

doi: https://doi.org/10.1016/S0166-4328(02)00384-4

ISSN 0166-4328. doi: https://doi.org/10.1016/S0166-4328(02)00384-4. URLhttps://www.sciencedirect.com/science/article/pii/S0166432802003844. Andreas Kalckert and H. Henrik Ehrsson. Moving a rubber hand that feels like your own: A dissociation of ownership and agency.Frontiers in Human Neuroscience, 6:40,

work page doi:10.1016/s0166-4328(02)00384-4
[9]

Diederik P Kingma and Jimmy Ba

doi: 10.3389/fnhum.2012.00040. Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980,

work page doi:10.3389/fnhum.2012.00040 2012
[10]

URL https://www

doi: 10.1126/scirobotics.aau9354. URL https://www. science.org/doi/abs/10.1126/scirobotics.aau9354. Pablo Lanillos, Jordi Pages, and Gordon Cheng. Robot self/other distinction: active inference meets neural networks learning in a mirror. InECAI 2020: 24th European Conference on Artificial Intelligence, 29 August–8 September 2020, Santiago de Compostela, S...

work page doi:10.1126/scirobotics.aau9354 2020
[11]

Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101,

Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101,

Pith/arXiv arXiv
[12]

Isaac gym: High performance gpu-based physics simulation for robot learning.arXiv preprint arXiv:2108.10470,

Viktor Makoviychuk, Lukasz Wawrzyniak, Yunrong Guo, Michelle Lu, Kier Storey, Miles Macklin, David Hoeller, Nikita Rudin, Arthur Allshire, Ankur Handa, et al. Isaac gym: High performance gpu-based physics simulation for robot learning.arXiv preprint arXiv:2108.10470,

Pith/arXiv arXiv
[13]

Srinivasan, Matthew Tancik, Jonathan T

Ben Mildenhall, Pratul P . Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ra- mamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view synthesis. InComputer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part I, pp. 405–421, Berlin, Heidelberg,

2020
[15]

URL https://doi.org/10.1007/978-3-030-58452-8

work page doi:10.1007/978-3-030-58452-8
[16]

Representation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748,

Aaron van den Oord, Yazhe Li, and Oriol Vinyals. Representation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748,

Pith/arXiv arXiv
[17]

Philippe Rochat and Tricia Striano

doi: 10.1037/0012-1649.31.4.626. Philippe Rochat and Tricia Striano. Perceived self in infancy.Infant behavior and development, 23(3-4):513–530,

work page doi:10.1037/0012-1649.31.4.626
[18]

Yokoyama, S

doi: 10.1109/ICRA57147.2024.10611047. Christina Soyoung Song and Youn-Kyung Kim. The role of the human-robot interaction in consumers’ acceptance of humanoid retail service robots.Journal of Business Research, 146: 489–503,

work page doi:10.1109/icra57147.2024.10611047 2024
[19]

doi: https://doi.org/10.1016/j.jbusres.2022.03.087

ISSN 0148-2963. doi: https://doi.org/10.1016/j.jbusres.2022.03.087. URL https://www.sciencedirect.com/science/article/pii/S014829632200323X. Nikolaus Steinbeis. The role of self–other distinction in understanding others’ mental and emotional states: neurocognitive mechanisms in children and adults.Philosophical Transactions of the Royal Society B: Biologi...

work page doi:10.1016/j.jbusres.2022.03.087 2022

[1] [1]

Retargeting matters: General motion retargeting for humanoid motion tracking.arXiv preprint arXiv:2510.02252,

Joao Pedro Araujo, Yanjie Ze, Pei Xu, Jiajun Wu, and C Karen Liu. Retargeting matters: General motion retargeting for humanoid motion tracking.arXiv preprint arXiv:2510.02252,

arXiv

[2] [2]

Josh Bongard, Victor Zykov, and Hod Lipson

doi: 10.1037/0012-1649.21.6.963. Josh Bongard, Victor Zykov, and Hod Lipson. Resilient machines through continuous self-modeling.Science, 314(5802):1118–1121,

work page doi:10.1037/0012-1649.21.6.963

[3] [3]

URL https://www.science.org/doi/abs/10.1126/science.1133687

doi: 10.1126/science.1133687. URL https://www.science.org/doi/abs/10.1126/science.1133687. Celia A. Brownell, Stephanie Zerwas, and Geetha B. Ramani. “So big”: The development of body self-awareness in toddlers.Child Development, 78(5):1426–1440,

work page doi:10.1126/science.1133687

[4] [4]

Tangney, Roy F

doi: 10.1111/j. 1467-8624.2007.01075.x. Nicolas Carion, Laura Gustafson, Yuan-Ting Hu, Shoubhik Debnath, Ronghang Hu, Didac Suris, Chaitanya Ryali, Kalyan Vasudev Alwala, Haitham Khedr, Andrew Huang, et al. Sam 3: Segment anything with concepts.arXiv preprint arXiv:2511.16719,

work page doi:10.1111/j 2007

[5] [5]

Shaun Gallagher

doi: 10.1016/j.tics.2003.10.004. Shaun Gallagher. Philosophical conceptions of the self: implications for cognitive science. Trends in cognitive sciences, 4(1):14–21,

work page doi:10.1016/j.tics.2003.10.004 2003

[6] [6]

Nerf: Neural radiance field in 3d vision, a comprehensive review.arXiv preprint arXiv:2210.00379,

Kyle Gao, Yina Gao, Hongjie He, Dening Lu, Linlin Xu, and Jonathan Li. Nerf: Neural radiance field in 3d vision, a comprehensive review.arXiv preprint arXiv:2210.00379,

Pith/arXiv arXiv

[7] [7]

17 Yuhang Hu, Boyuan Chen, and Hod Lipson

doi: 10.1109/TMECH.2025.3579247. 17 Yuhang Hu, Boyuan Chen, and Hod Lipson. Egocentric visual self-modeling for autonomous robot dynamics prediction and adaptation.npj Robotics, 3, 06 2025a. doi: 10.1038/ s44182-025-00031-6. Yuhang Hu, Jiong Lin, and Hod Lipson. Teaching robots to build simulations of themselves. Nature Machine Intelligence, 7(3):484–494,...

work page doi:10.1109/tmech.2025.3579247 2025

[8] [8]

doi: https://doi.org/10.1016/S0166-4328(02)00384-4

ISSN 0166-4328. doi: https://doi.org/10.1016/S0166-4328(02)00384-4. URLhttps://www.sciencedirect.com/science/article/pii/S0166432802003844. Andreas Kalckert and H. Henrik Ehrsson. Moving a rubber hand that feels like your own: A dissociation of ownership and agency.Frontiers in Human Neuroscience, 6:40,

work page doi:10.1016/s0166-4328(02)00384-4

[9] [9]

Diederik P Kingma and Jimmy Ba

doi: 10.3389/fnhum.2012.00040. Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980,

work page doi:10.3389/fnhum.2012.00040 2012

[10] [10]

URL https://www

doi: 10.1126/scirobotics.aau9354. URL https://www. science.org/doi/abs/10.1126/scirobotics.aau9354. Pablo Lanillos, Jordi Pages, and Gordon Cheng. Robot self/other distinction: active inference meets neural networks learning in a mirror. InECAI 2020: 24th European Conference on Artificial Intelligence, 29 August–8 September 2020, Santiago de Compostela, S...

work page doi:10.1126/scirobotics.aau9354 2020

[11] [11]

Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101,

Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization.arXiv preprint arXiv:1711.05101,

Pith/arXiv arXiv

[12] [12]

Isaac gym: High performance gpu-based physics simulation for robot learning.arXiv preprint arXiv:2108.10470,

Viktor Makoviychuk, Lukasz Wawrzyniak, Yunrong Guo, Michelle Lu, Kier Storey, Miles Macklin, David Hoeller, Nikita Rudin, Arthur Allshire, Ankur Handa, et al. Isaac gym: High performance gpu-based physics simulation for robot learning.arXiv preprint arXiv:2108.10470,

Pith/arXiv arXiv

[13] [13]

Srinivasan, Matthew Tancik, Jonathan T

Ben Mildenhall, Pratul P . Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ra- mamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view synthesis. InComputer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part I, pp. 405–421, Berlin, Heidelberg,

2020

[14] [15]

URL https://doi.org/10.1007/978-3-030-58452-8

work page doi:10.1007/978-3-030-58452-8

[15] [16]

Representation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748,

Aaron van den Oord, Yazhe Li, and Oriol Vinyals. Representation learning with contrastive predictive coding.arXiv preprint arXiv:1807.03748,

Pith/arXiv arXiv

[16] [17]

Philippe Rochat and Tricia Striano

doi: 10.1037/0012-1649.31.4.626. Philippe Rochat and Tricia Striano. Perceived self in infancy.Infant behavior and development, 23(3-4):513–530,

work page doi:10.1037/0012-1649.31.4.626

[17] [18]

Yokoyama, S

doi: 10.1109/ICRA57147.2024.10611047. Christina Soyoung Song and Youn-Kyung Kim. The role of the human-robot interaction in consumers’ acceptance of humanoid retail service robots.Journal of Business Research, 146: 489–503,

work page doi:10.1109/icra57147.2024.10611047 2024

[18] [19]

doi: https://doi.org/10.1016/j.jbusres.2022.03.087

ISSN 0148-2963. doi: https://doi.org/10.1016/j.jbusres.2022.03.087. URL https://www.sciencedirect.com/science/article/pii/S014829632200323X. Nikolaus Steinbeis. The role of self–other distinction in understanding others’ mental and emotional states: neurocognitive mechanisms in children and adults.Philosophical Transactions of the Royal Society B: Biologi...

work page doi:10.1016/j.jbusres.2022.03.087 2022