CoChat
Sign Up
Cited by 58Embodied and Extended CognitionReinforcement Learning in RoboticsAction Observation and Synchronization

End-to-End Pixel-Based Deep Active Inference for Body Perception and Action

Open access
Sancaktar, Cansu, van Gerven, Marcel A. J., Lanillos, Pablo·2020·2020 Joint IEEE 10th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)
DOI 10.1109/icdl-epirob48136.2020.9278105
Open this paper in CoChat — free accountOpen PDF

Abstract

We present a pixel-based deep active inference algorithm (PixelAI) inspired by human body perception and action. Our algorithm combines the free energy principle from neuroscience, rooted in variational inference, with deep convolutional decoders to scale the algorithm to directly deal with raw visual input and provide online adaptive inference. Our approach is validated by studying body perception and action in a simulated and a real Nao robot. Results show that our approach allows the robot to perform 1) dynamical body estimation of its arm using only monocular camera images and 2) autonomous reaching to “imagined” arm poses in visual space. This suggests that robot and human body perception and action can be efficiently solved by viewing both as an active inference problem guided by ongoing sensory input.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/icdl-epirob48136.2020.9278105

Cites

6 of 6 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The DAI_HR approaches cannot be considered complete implementations of deep active inference because they are restricted to a single decoder and lack encoder, transition, and critic networks.

    “Since, Sancaktar et al (2020); Lanillos et al (2020); Oliver et al (2019) restricted themself to the context of a single decoder, this approach can not be considered as a complete implementation of deep active inference.More precisely, the encoder, transition and critic (or policy) networks are missing.”

    From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Sancaktar et al (2020)

    Section 3.2 (p. 5) describes a convolutional decoder as the visual generative model, and Section 6 (p. 12) says actions are generated without learning an explicit policy. These passages support the architectural description, but the paper does not say that the absence of encoder, transition, and critic networks makes its approach an incomplete implementation of deep active inference. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • The DAI_HR papers propose a free energy minimization scheme based on a single decoder network used to control Nao, TIAGo, and iCub robots.

    “Briefly, those papers propose a free energy minimization scheme based on a single decoder network, which is used to control Nao, TIAGo, and iCub robots, respectively.”

    From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Sancaktar et al (2020)

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 24

  1. [1][1] Ruzena Bajcsy, Yiannis Aloimonos, and John K Tsotsos. Revisiting active perception. Autonomous Robots , 42(2):177-196, 2018. -
  2. [3][3] Christopher L Buckley, Chang Sub Kim, Simon McGregor, and Anil K Seth. The free energy principle for action and perception: A mathematical review. Journal of Mathematical Psychology , 2017. -
  3. [4][4] Cristina Garcia Cifuentes, Jan Issac, Manuel Wüthrich, Stefan Schaal, and Jeannette Bohg. Probabilistic articulated real-time tracking for robot manipulation. IEEE Robotics and Automation Letters , 2(2):577-584, 2016. -
  4. [5][5] Peter Dayan, Geoffrey E Hinton, Radford M Neal, and Richard S Zemel. The helmholtz machine. Neural computation , 7(5):889-904, 1995. -
  5. [6][6] German Diez-Valencia, Takuya Ohashi, Pablo Lanillos, and Gordon Cheng. Sensorimotor learning for artificial body perception. arXiv preprint arXiv:1901.09792 , 2019. - · arXiv
  6. [7][7] Alexey Dosovitskiy, Jost Tobias Springenberg, Maxim Tatarchenko, and Thomas Brox. Learning to generate chairs, tables and cars with convolutional networks. IEEE transactions on pattern analysis and machine intelligence , 39(4):692-705, 2016. -
  7. [8][8] Kenji Doya. What are the computations of the cerebellum, the basal ganglia and the cerebral cortex? Neural networks , 12(7-8):961-974, 1999. -
  8. [9][9] Vincent Dumoulin and Francesco Visin. A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:1603.07285 , 2016. - · arXiv
  9. [10][10] Karl Friston, Jérémie Mattout, Nelson Trujillo-Barreto, John Ashburner, and Will Penny. Variational free energy and the laplace approximation. Neuroimage , 34(1):220-234, 2007. -
  10. [12][12] David Gouaillier, Cyrille Collette, and Chris Kilner. Omni-directional closed-loop walk for nao. In 2010 10th IEEE-RAS International Conference on Humanoid Robots , pages 448-454. IEEE, 2010. -
  11. [14][14] Geoffrey E Hinton and Richard S Zemel. Autoencoders, minimum description length and helmholtz free energy. In Advances in neural information processing systems , pages 3-10, 1994. -
  12. [15][15] Nina-Alisa Hinz, Pablo Lanillos, Hermann Mueller, and Gordon Cheng. Drifting perceptual patterns suggest prediction errors fusion rather than hypothesis selection: replicating the rubberhand illusion on a robot. arXiv preprint arXiv:1806.06809 , 2018. - · arXiv
  13. [16][16] Seth Hutchinson, Gregory D Hager, and Peter I Corke. A tutorial on visual servo control. IEEE transactions on robotics and automation , 12(5):651-670, 1996. -
  14. [17][17] Diederik P Kingma and Max Welling. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114 , 2013. - · arXiv
  15. [18][18] Pablo Lanillos, Emmanuel Dean-Leon, and Gordon Cheng. Yielding self-perception in robots through sensorimotor contingencies. IEEE Trans. on Cognitive and Developmental Systems , (99):1-1, 2016. -
  16. [19][19] Sergey Levine, Chelsea Finn, Trevor Darrell, and Pieter Abbeel. End-to-end training of deep visuomotor policies. The Journal of Machine Learning Research , 17(1):1334-1373, 2016. -
  17. [20][20] Kingson Man and Antonio Damasio. Homeostasis and soft robotics in the design of feeling machines. Nature Machine Intelligence , 1(10):446-452, 2019. -
  18. [21][21] Beren Millidge. Deep active inference as variational policy gradients. arXiv preprint arXiv:1907.03876 , 2019. - · arXiv
  19. [22][22] Ashvin V Nair, Vitchyr Pong, Murtaza Dalal, Shikhar Bahl, Steven Lin, and Sergey Levine. Visual reinforcement learning with imagined goals. In Advances in Neural Information Processing Systems , pages 9191-9200, 2018. -
  20. [23][23] Augustus Odena, Vincent Dumoulin, and Chris Olah. Deconvolution and checkerboard artifacts. Distill , 1(10):e3, 2016. -
  21. [27][27] Marc Toussaint. Robot trajectory optimization using approximate inference. In Proceedings of the 26th annual international conference on machine learning , pages 1049-1056, 2009. -
  22. [28][28] Alexander Tschantz, Manuel Baltieri, Anil Seth, Christopher L Buckley, et al. Scaling active inference. arXiv preprint arXiv:1911.10601 , 2019. - · arXiv
  23. [29][29] Martin J Wainwright and Michael I Jordan. Graphical models, exponential families, and variational inference. Foundations and Trends in Machine Learning , 1(1):1-305, 2007. -
  24. [30][30] Daniel M Wolpert, Jörn Diedrichsen, and J Randall Flanagan. Principles of sensorimotor learning. Nature Reviews Neuroscience , 12(12):739, 2011. -

Cited by

Sorted by citations

Authors
Sancaktar, C., van Gerven, M. A. J., & Lanillos, P.
Affiliations
  • Technical University of Munich (DE)
  • Radboud University Nijmegen (NL)
ORCID
Lanillos, Pablo
Venue
2020 Joint IEEE 10th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)
Publisher
IEEE
Published
2020-10-26
Pages
1-8
DOI
10.1109/icdl-epirob48136.2020.9278105
OpenAlex
W2999994890
Cited by
58
References
6
Citations by year
2026: 1 · 2025: 7 · 2024: 6 · 2023: 10 · 2022: 8 · 2021: 17 · 2020: 9
FWCI
14.33
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE International Conference on Development and Learning (ICDL) (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Radboud Repository (Radboud University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Embodied and Extended Cognition
  • Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
  • Social Sciences › Psychology › Social Psychology › Action Observation and Synchronization
Keywords
free energy principle, body perception, body action, NAO robot, monocular vision, convolutional decoder, human-robot interaction

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review