1. Neuroscience

Unimodal statistical learning produces multimodal object-like representations

  1. Gábor Lengyel  Is a corresponding author
  2. Goda Žalalytė
  3. Alexandros Pantelides
  4. James Neilson Ingram
  5. József Fiser  Is a corresponding author
  6. Máté Lengyel  Is a corresponding author
  7. Daniel M Wolpert  Is a corresponding author
  1. Central European University, Hungary
  2. University of Cambridge, United Kingdom
https://doi.org/10.7554/eLife.43942
Share this article
Cite this article
  1. Gábor Lengyel
  2. Goda Žalalytė
  3. Alexandros Pantelides
  4. James Neilson Ingram
  5. József Fiser
  6. Máté Lengyel
  7. Daniel M Wolpert
(2019)
Unimodal statistical learning produces multimodal object-like representations
eLife 8:e43942.
https://doi.org/10.7554/eLife.43942
  2. Download BibTeX
  3. Download .RIS

Abstract

The concept of objects is fundamental to cognition and is defined by a consistent set of sensory properties and physical affordances. Although it is unknown how the abstract concept of an object emerges, most accounts assume that visual or haptic boundaries are crucial in this process. Here, we tested an alternative hypothesis that boundaries are not essential but simply reflect a more fundamental principle: consistent visual or haptic statistical properties. Using a novel visuo-haptic statistical learning paradigm, we familiarised participants with objects defined solely by across-scene statistics provided either visually or through physical interactions. We then tested them on both a visual familiarity and a haptic pulling task, thus measuring both within-modality learning and across-modality generalisation. Participants showed strong within-modality learning and 'zero-shot' across-modality generalisation which were highly correlated. Our results demonstrate that humans can segment scenes into objects, without any explicit boundary cues, using purely statistical information.

Data availability

The scripts for all of the analysis reported in the manuscript can be found here https://github.com/GaborLengyel/Visual-Haptic-Statistical-Learning. There is a README file that explains both where the data can be found (Open Science Framework https://osf.io/456qb/) and how to run the analysis.

Article and author information

Author details

  1. Gábor Lengyel

    Department of Cognitive Science, Central European University, Budapest, Hungary
    For correspondence
    lengyel.gaabor@gmail.com
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1535-3250

  2. Goda Žalalytė

    Computational and Biological Learning Lab, Department of Engineering, University of Cambridge, Cambridge, United Kingdom
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0012-9950

  3. Alexandros Pantelides

    Computational and Biological Learning Lab, Department of Engineering, University of Cambridge, Cambridge, United Kingdom
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6234-6061

  4. James Neilson Ingram

    Computational and Biological Learning Lab, Department of Engineering, University of Cambridge, Cambridge, United Kingdom
    Competing interests
    The authors declare that no competing interests exist.

  5. József Fiser

    Department of Cognitive Science, Central European University, Budapest, Hungary
    For correspondence
    fiserj@ceu.edu
    Competing interests
    The authors declare that no competing interests exist.

  6. Máté Lengyel

    Computational and Biological Learning Lab, Department of Engineering, University of Cambridge, Cambridge, United Kingdom
    For correspondence
    m.lengyel@eng.cam.ac.uk
    Competing interests
    The authors declare that no competing interests exist.

  7. Daniel M Wolpert

    Computational and Biological Learning Lab, Department of Engineering, University of Cambridge, Cambridge, United Kingdom
    For correspondence
    wolpert@eng.cam.ac.uk
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-2011-2790

Funding

ERC Consolidator Grant (ERC-2016-COG/726090)

  • Máté Lengyel

Royal Society Noreen Murray Professorship in Neurobiolog (RP120142)

  • Daniel M Wolpert

EU-FP7 Marie Curie CIG (CIG 618918)

  • József Fiser

Wellcome Trust: New Investigator Award (095621/Z/11/Z)

  • Máté Lengyel

National Institutes of Health (NIH R21 HD088731)

  • József Fiser

Wellcome Trust: Senior Investigator Award (097803/Z/11/Z)

  • Daniel M Wolpert

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Ethics

Human subjects: All participants gave informed consent. All experimental protocols were approved by the University of Cambridge Psychology Ethics Committee.

Copyright

© 2019, Lengyel et al.

This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 2,583
    views
  • 416
    downloads
  • 18
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Gábor Lengyel
  2. Goda Žalalytė
  3. Alexandros Pantelides
  4. James Neilson Ingram
  5. József Fiser
  6. Máté Lengyel
  7. Daniel M Wolpert
(2019)
Unimodal statistical learning produces multimodal object-like representations
eLife 8:e43942.
https://doi.org/10.7554/eLife.43942

Share this article

https://doi.org/10.7554/eLife.43942

Categories and tags

Research organism

Further reading

    1. Neuroscience

    Rab10 regulates neuropeptide release by maintaining Ca2+ homeostasis and protein synthesis

    Jian Dong, Mian Chen ... Matthijs Verhage
    Research Article

    Dense core vesicles (DCVs) transport and release various neuropeptides and neurotrophins that control diverse brain functions, but the DCV secretory pathway remains poorly understood. Here, we tested a prediction emerging from invertebrate studies about the crucial role of the intracellular trafficking GTPase Rab10, by assessing DCV exocytosis at single-cell resolution upon acute Rab10 depletion in mature mouse hippocampal neurons, to circumvent potential confounding effects of Rab10’s established role in neurite outgrowth. We observed a significant inhibition of DCV exocytosis in Rab10-depleted neurons, whereas synaptic vesicle exocytosis was unaffected. However, rather than a direct involvement in DCV trafficking, this effect was attributed to two ER-dependent processes, ER-regulated intracellular Ca2+ dynamics, and protein synthesis. Gene Ontology analysis of differentially expressed proteins upon Rab10 depletion identified substantial alterations in synaptic and ER/ribosomal proteins, including the Ca2+ pump SERCA2. In addition, ER morphology and dynamics were altered, ER Ca2+ levels were depleted, and Ca2+ homeostasis was impaired in Rab10-depleted neurons. However, Ca2+ entry using a Ca2+ ionophore still triggered less DCV exocytosis. Instead, leucine supplementation, which enhances protein synthesis, largely rescued DCV exocytosis deficiency. We conclude that Rab10 is required for neuropeptide release by maintaining Ca2+ dynamics and regulating protein synthesis. Furthermore, DCV exocytosis appeared more dependent on (acute) protein synthesis than synaptic vesicle exocytosis.

    1. Neuroscience

    Neuronal Signaling: At the crossroads of calcium signaling, protein synthesis and neuropeptide release

    Jakob Rupert, Dragomir Milovanovic
    Insight

    By influencing calcium homeostasis, local protein synthesis and the endoplasmic reticulum, a small protein called Rab10 emerges as a crucial cytoplasmic regulator of neuropeptide secretion.

    1. Neuroscience

    Peripheral opioid receptor antagonism alleviates fentanyl-induced cardiorespiratory depression and is devoid of aversive behavior

    Brian C Ruyle, Sarah Masud ... Jose A Morón
    Research Article

    Millions of Americans suffering from Opioid Use Disorders face a high risk of fatal overdose due to opioid-induced respiratory depression (OIRD). Fentanyl, a powerful synthetic opioid, is a major contributor to the rising rates of overdose deaths. Reversing fentanyl overdoses has proved challenging due to its high potency and the rapid onset of OIRD. We assessed the contributions of central and peripheral mu opioid receptors (MORs) in mediating fentanyl-induced physiological responses. The peripherally restricted MOR antagonist naloxone methiodide (NLXM) both prevented and reversed OIRD to a degree comparable to that of naloxone (NLX), indicating substantial involvement of peripheral MORs to OIRD. Interestingly, NLXM-mediated OIRD reversal did not produce aversive behaviors observed after NLX. We show that neurons in the nucleus of the solitary tract (nTS), the first central synapse of peripheral afferents, exhibit a biphasic activity profile following fentanyl exposure. NLXM pretreatment attenuates this activity, suggesting that these responses are mediated by peripheral MORs. Together, these findings establish a critical role for peripheral MORs, including ascending inputs to the nTS, as sites of dysfunction during OIRD. Furthermore, selective peripheral MOR antagonism could be a promising therapeutic strategy for managing OIRD by sparing CNS-driven acute opioid-associated withdrawal and aversion observed after NLX.