Unsupervised Natural Experience Rapidly Alters Invariant Object Representation in Visual Cortex

doi:10.1126/science.1160028

Paper

Unsupervised Natural Experience Rapidly Alters Invariant Object Representation in Visual Cortex

Published Sep 12, 2008 · Nuo Li, J. DiCarlo

Science

Q1 SJR score

90

Citations

5

Influential Citations

Full text Semantic Scholar

Abstract

Object recognition is challenging because each object produces myriad retinal images. Responses of neurons from the inferior temporal cortex (IT) are selective to different objects, yet tolerant (“invariant”) to changes in object position, scale, and pose. How does the brain construct this neuronal tolerance? We report a form of neuronal learning that suggests the underlying solution. Targeted alteration of the natural temporal contiguity of visual experience caused specific changes in IT position tolerance. This unsupervised temporal slowness learning (UTL) was substantial, increased with experience, and was significant in single IT neurons after just 1 hour. Together with previous theoretical work and human object perception experiments, we speculate that UTL may reflect the mechanism by which the visual stream builds and maintains tolerant object representations.

Highly Cited

Study Snapshot

Key takeawayUnsupervised temporal slowness learning (UTL) rapidly alters invariant object representation in the visual cortex, potentially reflecting the mechanism by which the visual stream builds and maintains tolerant object representations.

PopulationOlder adults (50-71 years)

Sample size24

MethodsObservational

OutcomesBody Mass Index projections

ResultsSocial networks mitigate obesity in older groups.

Sign up to use Study Snapshot

Consensus is limited without an account. Create an account or sign in to get more searches and use the Study Snapshot.

Create an account

Paper

Unsupervised Natural Experience Rapidly Alters Invariant Object Representation in Visual Cortex

Sign up to use Study Snapshot

References

Slowness: An Objective for Spike-Timing–Dependent Plasticity?

Spike-timing-dependent plasticity (STDP) can implement slowness learning in linear Poisson neurons, enabling efficient object recognition despite changes in context.

Unsupervised Learning of Visual Features through Spike Timing Dependent Plasticity

Spike timing dependent plasticity (STDP) in a spiking neural network can lead to fast and selective learning of intermediate-complexity visual features, enabling robust object recognition in natural images.

A Model of the Ventral Visual System Based on Temporal Stability and Local Memory

The ventral visual pathway can be modelled using optimal stability of sensory representations and local memory, suggesting that functionally heterogeneous cortical areas can be generated by a few computational principles.

Learning and neural plasticity in visual object recognition

The adult brain uses adaptive coding strategies and neural plasticity to rapidly and accurately recognize coherent visual objects in cluttered scenes, surpassing artificial vision systems.

Object Selectivity of Local Field Potentials and Spikes in the Macaque Inferior Temporal Cortex

Local field potentials in the monkey inferior temporal cortex show strong object selectivity, with inputs to IT neurons conveying selectivity for complex shapes and spanning several millimeters.

Citations

Digital Twin Studies for Reverse Engineering the Origins of Visual Intelligence.

Digital twin studies reveal that both empiricist and nativist theories about intelligence can be explained by a universal principle called space-time fitting, which explains learning across humans, animals, and machines.

Identifying dyslexia in school pupils from eye movement and demographic data using artificial intelligence

Combining eye-movement data and demographic data using artificial intelligence models can effectively identify dyslexia in primary school pupils, with IQ, gender, and age being the most important features.

Static and dynamic resting-state brain activity patterns of table tennis players in 7-Tesla MRI

Table tennis players exhibit altered visuomotor transformation and executive function pathways, with the left inferior temporal gyrus playing a vital role in both pathways.

The combination of Hebbian and predictive plasticity learns invariant object representations in deep sensory networks

Latent Predictive Learning (LPL) combines Hebbian and predictive plasticity to form invariant object representations in deep sensory networks, enabling automatic object discrimination in response to sensory stimuli.

Local minimization of prediction errors drives learning of invariant object representations in a generative network model of visual perception

A predictive coding network can learn both bottom-up and top-down visual perception by minimizing prediction errors, generating invariant object representations, and reconstructing whole objects from partially occluded input images.